Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons from a FinTech crisis

Stefan Zeranski; Ibrahim E. Sancak

doi:10.1057/s41310-021-00111-7

Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons from a FinTech crisis

Zeranski, Stefan; Sancak, Ibrahim E. 2021-12-01 00:00:00 The U.S. financial markets faced an unprecedented rapid decline and recovery on May 6, 2010, known as the May 6 flash crash. Roughly one trillion $ market value in less than thirty minutes vanished with the biggest one-day point decline in the history of the DJIA at the time. Since the market events took place in electronic markets, and algorithmic trading and high- frequency trading, parts of FinTech, played significant roles, we handle the May 6 flash crash from the FinTech, SupTech, and financial supervision perspectives. With the flashback method, we analyzed the reactions of market participants, media, and two financial supervisors, the SEC, and the CFTC, to the market crash. We find that the technological imbalance between financial markets or institutions and their supervisors drove the markets in uncertainty, hence in a fear and panic environment. Since the imbalance has not diminished yet, the same risks still exist. As a remedy, we introduce a new concept and model with a well-functioning SupTech system to cope with the May 6 type FinTech crises. Keywords Supervisory technology · SupTech · FinTech · RegTech · Financial supervision · Financial system · Prudential supervisory disclosure · Financial authority · Digital finance · May 6 flash crash · FinTech crises · Financial crises · Financial stability · Informational efficiency · Systemically important data · Systemically important markets · Know-your- data · Know-your-technology · Know-your-markets · Inform-your-markets JEL Classification D47 · D53 · G18 · G01 · G28 · H11 · K22 · K23 · L15 · O31 · O32 Abbreviations CTS Consolidated Tape System AI Artificial Intelligence DJIA Dow Jones Industrial Average AT Algorithmic Trading EBA European Banking Authority BaFin German Federal Financial Supervi-EC European Commission sory Authority ECB European Central Bank BCBS Basel Committee on Banking EDGAR U.S. Electronic Data Gathering, Supervision Analysis, and Retrieval System BIS Bank for International Settlements EFIF European Forum for Innovation CAT Consolidated Audit Trail Facilitators CMB Capital Markets Board of Turkey EIOPA European Insurance and Occupa- CFTC U.S. Commodity Futures Trading tional Pensions Authority Commission E-Mini S&P 500 Futures Contracts CME Chicago Mercantile Exchange ESMA European Securities and Markets CQS Consolidated Quotation System Authority ETF Exchange Traded Fund EU European Union * Ibrahim E. Sancak EWS Early Warning System i.sancak@ostfalia.de FED U.S. Federal Reserve System Stefan Zeranski FDIC Federal Deposit Insurance st.zeranski@ostfalia.de Corporation Ostfalia University of Applied Sciences, Wolfenbüttel, Germany Vol.:(0123456789) 316 S. Zeranski, I. E. Sancak FINRA U.S. Financial Industry Regulatory Since the October 1987 crash originated at the U.S. mar- Authority kets, the financial market structure has evolved as techno- FinTech Financial Technology logical advancements that have enabled participants to trade FSB Financial Stability Board using algorithms with little or no human intervention (Kir- FSP Financial Services Provider ilenko et al. 2018). Today, digital finance and the FinTech HFT High-frequency Trading world have a lot of tools and technologies, including high- HFTs High-frequency Traders frequency trading (HFT) and algorithmic trading (AT). HFT IMF International Monetary Fund and AT were intensively debated due to the May 6, 2010, LRPs Liquidity Replenishment Points flash crash and received close attention by the public and MIDAS Market Information Data Analytics regulators (Gomber et al. 2017). This debate also triggered System intensive academic research on the impact of high-frequency NBBO National Best Bid and Offer trading and algorithmic trading on market quality, especially NLP Natural Language Processing market stability and integrity (Gomber et al. 2017). How- NMS National Market System ever, the roles of supervisors in the digital financial world NYSE New York Stock Exchange have not been discussed enough to develop new instruments OCC Office of the Comptroller of the to answer new risks arising from new technological tools, Currency market speed, and big data. The May 6 flash crash is a con- PSD Prudential Supervisory Disclosure venient case to analyze the roles and possible responding RegTech Regulatory Technology technologies and policies against FinTech-related risks. SEC U.S. Securities and Exchange The market events of May 6, 2010, shook the confidence Commission of market participants and raised questions about the mar- SPY S&P 500 SPDR Exchange Traded ket structure of electronic markets (Kirilenko et al. 2017). Fund (SPDR: Standard and Poor’s Considering the regulatory and supervisory responsibilities, Depository Receipt) it also raised questions about possible responses with SupT- SRO Self-Regulatory Organization ech. Most academic papers and discussions about the May SSM Single Supervisory Mechanism 6 flash crash have focused on the market microstructure. SupTech Supervisory Technology However, supervisory agencies’ roles are not less impor- TECHs in Finance FinTech, RegTech, SupTech, and tant than that. We believe that the May 6 case has more other “Tech” areas in Finance and at issues and implications than the concerns about the market Financial Sectors. structure. For example, the May 6 case addresses a market U.S. United States of America disorder that nurtures fear and panic and feeds market-made VIX Volatility Index detrimental stories about the crash, which might fuel further VPIN Volume-Synchronized Probability of crashes and vicious circles, ultimately crises. Informed Trading One of the concerns about the May 6 case was uncertainty during the market crash. With a flashback perspective, we collected information about responses of market participants Introduction and media to the crash. No one, including the supervisors, did know what happened during the day, even after the day Technological imbalance or asymmetric technology between until the supervisory agencies revealed a joint report with financial markets or institutions and their supervisors is convincing findings, approximately five months later, on more dangerous than cyber-attack risks since cyberattacks September 30, 2010. are well-known risk types; hence, there is considerable vigi- The SEC’s chairperson announced with a written state- lance to develop shields against them. However, the lack ment on May 20, 2010, that: "On the Monday following the of a well-functioning supervisory technology (SupTech) events of May 6, I met here in Washington with the lead- leaves many doors wide-open for detrimental technologi- ers of six markets—New York Stock Exchange, NASDAQ cal transactions and their ensuing effects on an economy’s Stock Market, BATS Exchange, Direct Edge ECN, Inter- financial stability. In other words, asymmetric technology national Securities Exchange, and Chicago Board Options between financial markets and the relevant supervisors is Exchange—and FINRA, to discuss the causes of market one of the most significant risks today. Therefore, having a events of May 6, the potential contributing factors, and pos- digital financial supervisory system with a well-functioning sible market reforms. The meeting was productive and col- SupTech is one of the best risk management strategies in laborative, and there was a strong consensus that the type this regard. of aberrational volatility experienced on May 6 is wholly Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 317 Fig. 1 Decline of E-Mini and DJIA Based on 11:00 AM Lev- els. Source: Schapiro (2010); Bloomberg unacceptable in our markets." The May 6 case has forced debt crisis, led to growing uncertainty in the financial mar - the U.S. financial supervisors to reform their infrastructure. kets (CFTC and SEC 2010). This negative sentiment-driven Analyzing the May 6 case and examining the reform sell pressure and flight to quality transactions accelerated the efforts, we have realized that financial supervisors need overall decline in the financial markets suddenly beginning entirely new instruments to monitor and supervise today’s shortly after 2:30 PM (CFTC and SEC 2010). markets. This paper mainly focuses on prudential supervi- The following graph indicates the path and the timing of sory disclosure and the reasoning behind it with a real and the E-Mini and the DJIA movements on May 6, based on stunning case, the May 6 flash crash. 11:00 AM levels (Fig. 1). Prudential supervisory disclosure is the set of disclosure As shown by the graph, the market crash deepened in rules for supervisors to inform the market participants timely 30 min, but market indicators came back in one hour, with about market-wide harmful conditions and activities to pre- around 2% decline based on 11:00 AM levels. serve market integrity and protect markets from detrimen- The following table is the flashback of the events of May tal rumors, orders and transactions by using their SupTech 6 (Table 1). capacity. This paper’s concept is much more related to the On May 6, in the four-and-one-half minutes (from 2:41 biggest economies, such as the U.S. and the EU, since their p.m. through 2:45:27 PM), prices of the E-Mini had fallen state-based or member-based economic areas are more con- by more than 5%, and prices of the SPY suffered a decline ducive to the PSD model. of over 6% (SEC and CFTC 2010). Internal stabilizers or This paper consists of two sections. In the first section circuit breakers carried out an important function to pre- of this paper, we analyze the May 6 market crash from the vent further liquidity and price collapses. For example, at supervisory technology perspective and the media and mar- 2:45:28 PM, trading on the E-Mini was paused for five sec- ket participants’ reactions. The second section explains the onds when the Chicago Mercantile Exchange (CME) Stop general framework of the new concept, prudential supervi- Logic Functionality was triggered; hence, the sell-side pres- sory disclosure, and its implications for banking and capital sure in the E-Mini was partly alleviated, and the buy-side markets sectors. The conclusion gives a summary of our interest increased, and when trading resumed at 2:45:33 PM, findings, views, and highlights. prices stabilized and shortly after that, the E-Mini began to recover, followed by the SPY (SEC and CFTC 2010). We Analysis of the May 6 market crash Stop Logic Functionality pauses trading when the trading engine and aftermaths recognizes that it has a series of resting stop orders that could lead to a cascade and move the market up or down beyond a specified amount (SEC and CFTC 2010). Chronological order of the May 6 market crash In contrast to the fact that the Chicago Mercantile Exchange (CME) Stop Logic Functionality was triggered and functioned, the SEC and On May 6, 2010, the negative sentiment, unsettling political the CFTC (2010) reported that the staffs of the CFTC and SEC were and economic news from overseas concerning the European working together with the markets to consider recalibrating the exist- 318 S. Zeranski, I. E. Sancak Table 1 The May 6 Market Crash Timeline Phases Time Event Effects and Results Phase I 1:00 PM Due to the news from overseas concerning the The number of volatility pauses, also known as European debt crisis, broadly negative market Liquidity Replenishment Points (LRPs), trig- sentiment was affecting an increase in the price gered on the New York Stock Exchange (NYSE) volatility of some individual securities in individual equities listed and traded on that exchange began to substantially increase above average levels 2:30 PM The S&P 500 volatility index (VIX) was up 22.5 Yields of ten-year Treasuries fell as investors percent from the opening level engaged in a “flight to quality,” and selling pressure had pushed the Dow Jones Industrial Average (DJIA) down about 2.5% During the first phase, from the open through about 2:32 PM, prices were broadly declining across markets, with stock market index products sustaining losses of about 3% Phase 2:32 PM A large fundamental trader (a mutual fund com- Sell pressure was initially absorbed by: II plex) initiated a sell program to sell a total of high-frequency traders (HFTs) and other interme- 75,000 E-Mini contracts (valued at approxi- diaries in the futures market, mately $4.1 billion) as a hedge to an existing fundamental buyers in the futures market, and equity position via an automated execution cross-market arbitrageurs who transferred this sell algorithm and executed the sell program by pressure to the equities markets by opportun- only targeting trading volume, and neither price istically buying E-Mini contracts and simul- nor time, extremely rapidly in just 20 min taneously selling products like SPY or selling individual equities in the S&P 500 Index As a result, HFTs accumulated a net long position of about 3,300 contracts From about 2:32 PM through about 2:41 PM, the broad markets began to lose more ground Phase 2:41 PM- 2:44 PM HFTs aggressively sold about 2,000 E-Mini con- Two liquidity crises – one at the broad index III tracts to reduce their temporary long positions. level in the E-Mini, the other with respect to At the same time, HFTs traded nearly 140,000 individual stocks E-Mini contracts or over 33% of the total trad- ing volume 2:45:28 PM Trading on the E-Mini was paused for five sec- Sell-side pressure in the E-Mini was partly allevi- onds when the Chicago Mercantile Exchange ated and buy-side interest increased Stop Logic Functionality was triggered to prevent a cascade of further price declines Volume spiked upwards and the broad markets plummeted a further 5–6% to reach intra-day lows of 9–10% Phase 2:45:33 PM-3:00 PM Trading resumed Prices stabilized and shortly thereafter, the E-Mini IV began to recover, followed by the SPY Broad market indices recovered while at the same time many individual securities and ETFs experienced extreme price fluctuations and traded in a disorderly fashion Phase 3:00 PM -Closings Most securities had reverted to trading at prices After the market closed, the exchanges and FINRA V reflecting true consensus values met and jointly agreed to cancel (or break) all such trades under their respective “clearly erro- neous” trade rules Prices of most individual securities significantly recovered, and trading resumed in a more orderly fashion Source CFTC and SEC (May 2010) and SEC and CFTC (2010); outlined by the Authors *Within the NYSE’s hybrid floor/electronic trading model, on May 6, the NYSE implemented price-bands known as “liquidity replenishment points.”, or LRPs. LRPs are intended to act as a “speed bump” and to dampen volatility in a given stock by temporarily converting from an auto- mated market to a manual auction market when a price movement of sufficient size is reached SEC and CFTC (2010) In the course of the day, VIX, a measure of the expected volatility of the S&P 500 Index, increased by 31.7 percent, which was the fourth largest single day increase in VIX SEC and CFTC (2010) It was realized that especially in times of significant volatility, high trading volume is not necessarily a reliable indicator of market liquidity SEC and CFTC (2010) contemplate that the circuit breakers played an important though the circuit breakers at different markets function with role in alleviating the crash stress and adverse effects. Even different parameters and protect the markets from broader and higher price and liquidity collapses, they are mainly not for public disclosure and an information source against Footnote 2 (continued) market rumors and uncertainty. Circuit breakers might be a ing market-wide circuit breakers—none of which were triggered on window for fresh air and an opportunity to gain some time May 6—that apply across all equity trading venues and the futures for making a prudential supervisory disclosure, but they markets. Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 319 cannot carry out any supervisory disclosure role. Even, cir- Our idea is that only exchanges or SROs like FINRA and cuit breakers might signal unintended messages to markets financial supervisors, such as the SEC and the CFTC, could and market participants. Moreover, there is no guarantee help protect the markets by making a statement from further that circuit breakers work properly or work at all in every deterioration. Uncertainty and the fear about the causes of case. During the U.S. Treasury “Flash Rally”, on October the May 6 crash harmed the market integrity and market 15, 2014, market safeguards could not prevent large price quality. The harm could be less with the supervisors’ imme- movements (Bouveret et al. 2015). diate disclosure statement in the sense of prudential super- Cross-market propagation issues were also in the center visory disclosure. However, at the time, neither the CFTC of the May 6 market crash. Since many products and mar- nor the SEC had the capacity to inform the markets with the kets are connected to each other, a market crash may trigger facts about the crash. another one, multiple exchanges and trading platforms might One of the observations is that many (though not all) be affected without knowing the real causes of the crashes firms significantly curtailed or completely halted their trad- they face at their markets. For example, a stock market crash ing activities at some point during the afternoon of May at an exchange might trigger other crashes at derivatives 6 (SEC and CFTC 2010). Data integrity issues were their exchanges due to the underlying stocks or stock indices. In number one concern. This also addresses the prudential this case, the derivatives exchange members or managers supervisory disclosure. With the real-time data collection cannot know the real causes of the market crash at deriva- capacity from multiple data centers and having data integ- tives markets. However, a supervisory agency that has a real- rity control capacity, during the cases like the May 6 market time data collection capacity from all markets, either stock crash, the supervisors can inform the markets and eliminate markets or derivatives markets, can capture the causes of potential concerns or clarify the situation for better market a market crash or intervene with before a crash comes out quality. by applying advanced data analytics. Thus, the May 6 type Data-integrity pauses were the reality of markets at the FinTech crises only can be managed by a real-time market- time. There were some other concerns about this unusual wide data collection capacity, which is one of the essential market behaviors. For example, at the time, it could be features of a well-functioning SupTech system. Moreover, hypothesized that these delays were due to a manipulative we should point out that the supervisory model of the U.S. practice called “quote stuffing” in which high volumes of financial markets is still not functional, since considering quotes were purposely sent to exchanges in order to create the case above, the SEC and the CFTC have different market data delays that would afford the firm sending these quotes responsibilities. Without perfect coordination and collabora- a trading advantage (SEC and CFTC 2010). tion, which is not a case in many times among national finan- Even though neither FinTech nor SupTech was a part of cial supervisors, cross-market propagation issues cannot be financial terminology at the time, the SEC and the CFTC’s managed, even with the real-time data collection capacity. report (September 2010) addresses FinTech, RegTech, and The May 6 market crash indicated that during the crash, SupTech concepts (TECHs in Finance) many times without not only media and individual investors but also institu- using these terms. The following paragraph is one of the tional investors did not know about the real causes of the key statements indicating agencies’ concerns in this regard: crash since they were not able to see the whole picture of “The events of May 6 clearly demonstrate the importance the markets. Considering asymmetric information envi- of data in today’s world of fully-automated trading strate- ronment during the market crash, based on their respec- gies and systems. The SEC staff will therefore be working tive individual risk assessments, some market makers and closely with the market centers to help ensure the integrity other liquidity providers widened their quote spreads, others and reliability of their data processes, especially those that reduced oe ff red liquidity, and a signic fi ant number withdrew involve the publication of trades and quotes to the consoli- completely from the markets (SEC and CFTC 2010). Not dated tape. In addition, the SEC staff will be working with only did some withdraw, but arguably they became liquidity the market centers in exploring their members’ trading prac- consumers by dumping their inventories, thus exacerbating tices to identify any unintentional or potentially abusive or the crash (Easley et al. 2011). The flash crash might have manipulative conduct that may cause such system delays that been avoided, or at least tempered, had liquidity providers remained in the marketplace (Easley et al. 2011). The paper of Easley et al. (2011) suggests the Volume-Synchro- nized Probability of Informed Trading (VPIN) as a solution. The VPIN might capture the increasing toxicity of the order flow in the Even though the market safeguards, circuit breakers, were not trig- hours and days prior. The VPIN contract might be used with pru- gered on October 15, 2014, they were triggered previously (Bouveret dential supervisory disclosure to monitor and manage similar risks et al. 2015). dynamically. 320 S. Zeranski, I. E. Sancak inhibit the ability of market participants to engage in a fair cannot completely rule out these possibilities.” This state- and orderly process of price discovery.” ment and the preliminary report’s general findings suggest Being supervisors of the foremost capital markets in the that both agencies did not have enough evidence to rule out world, the SEC and the CFTC have been aware of the impor- rumors and detrimental news feeds. The SEC’s chairperson tance of technology. However, digital transformation with also touched on the “fat finger” rumors as follows (Schap- a cutting edge SupTech system and a country-wide mar- iro, Testimony Concerning the Severe Market Disruption on ket reform requires political leadership and strong financial May 6, 2010, 11 May 2010): “There have been reports in the support, which have been not entirely in the hands of the press about a “fat-finger” error where, it is hypothesized, agencies. The same bottleneck holds for almost all financial an order of billions of shares was entered, rather than an regulators and supervisors. intended order of millions of shares. While we cannot yet definitely rule that possibility out, neither our review nor Reactions of the market participants and the media reviews by the relevant exchanges and market participants have uncovered such an error.” The reactions of market participants during the day on May The May 6 market crash also attracted the attention of the 6 have been analyzed by both financial agencies and academ - media. Economic and financial media are important sources ics. The SEC and CFTC reports tried to capture the behav- of information for financial consumers. Peress (2014) dem- iors of different market actors. In response to the increased onstrates that the media influence the stock market by risk perceptions, some market makers and other liquid- increasing the speed with which information diffuses across ity providers widened their quote spreads, others reduced investors and is impounded into stock prices. offered liquidity, and a significant number withdrew com- Borch (2017) evaluates the experimental impact (the pletely from the markets (SEC and CFTC 2010). The inves- Flash Crash’s effect on market participants), the real eco- tigations of the staffs of the SEC and the CFTC revealed that nomic impact (the actual economic effects) and the potential the largest and most erratic price moves observed on May 6 systemic impact (the systemic risk involved in algorithmic were caused by withdrawals of liquidity and the subsequent trading, as illustrated in the crash), and subsequently argues execution of trades at stub quotes. that each impact is contestable. Academic papers also analyzed the market structure and Borch (2017), inter alia, discusses the following three reached some results for the FinTech world. For example, aspects of the May 6 flash crash for today’s financial mar - the papers of Kirilenko, Kyle, Samadi, & Tuzun, The Flash kets: (a) as an event that significantly changed how market Crash: High-Frequency Trading in an Electronic Market participants perceive markets; (b) as an event that generated (2017) and Automation, Intermediation and the Flash Crash a massive loss of value, and hence had or could have had (2018) assert that HFTs behave differently than traditional considerable economic effects; and/or (c) as an event that is market makers; their behavior is empirically more consistent symptomatic of a novel set of systemic risks associated with with quote sniping than traditional market-making. algorithmic finance. The final report of the May 6 case was published on Sep- We took snapshots of some available news feeds about tember 30, 2010, and it has more robust evidence about the the a fl sh crash to conceptualize the ee ff cts of the uninformed causes of the events. However, with the preliminary report, media risk (Table 2). the CFTC and SEC’s staffs were considering some working The executive vice president and head of operations at the hypotheses. There were no clear and definite findings even NYSE Euronext’s New York Stock Exchange commented on though the agencies delivered 150-page preliminary reports the May 6 crash that “This highlights the risks of electronic on May 18, 2010, twelve days later. trading. When you have low volatility, electronic trading For the informational efficiency and market integrity works very well. But there are risks. It highlights the need for concerns, market participants should be informed by the human-based intervention.” (Lauricella and McKay 2010). responsible authorities in critical times and should not be Four years later, a news portal commented on the case as left in the hands of rumor feeds. We also believe that mar- follows (CNBC 2014): ket participants should be well informed to reduce adverse The Dow Jones Industrial Average slumped nearly selection risks and other outcomes of the inefficient infor - 1000 points in a matter of minutes in the flash crash of mational environment. 2010, sending traders into a panic and inciting scrutiny Some financial news sources were mentioning about a of the U.S. equities markets that’s still being felt four “fat finger” issue as the triggering source of the crash. The years later. preliminary report’s response to these rumors and news The May 6, 2010, crash was initially blamed on a “fat- was as follows (CFTC and SEC 2010): “We have found no finger” error made at Citigroup-a theory that was later evidence that these events were triggered by “fat finger” shot down and ultimately attributed to investment firm errors, computer hacking, or terrorist activity, although we Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 321 Table 2 Economic and Financial Media Comments on the Market Events of May 6, 2010 News source Date Comment Reuters 6 May 2010 “The Dow suffered its biggest ever intraday point drop —998.5 points. The market’s fall may have been exacerbated by erroneous trades that showed some shares briefly fell to nearly zero The situation remained unclear long after the closing bell as the Nasdaq Stock Market and others said they would cancel multiple erroneous trades. Other exchanges scrambled to examine orders.” (Krudy 2010) Forbes.com 6 May 2010 “Why the market plunged so much and so fast in the middle of the afternoon isn’t entirely clear. Some blame an erroneous quote on Procter & Gamble (PG), saying it caused panic selling across the board. Others say the selloff was caused by a trading error on the Nasdaq. This so-called fat-finger error occurred when a trader accidentally entered an order to sell a billion shares rather than a million shares. Still others blame the rioting in Greece for the selloff. That rioting was widely broadcast on trading floors." (Janjigian, 2010) CNBC.com (TV) 6 May 2010 A commentator on TV: “…machines broke down…”, “…the system obviously broke down…” “…it broke down, machines broke down…” A speaker on TV: “…there should be an investigation…” A speaker asks a question about the P&G prices: “…with P&G is there any rational way that you could describe P&G being three percent down and then 25 within what was in 90 s to three minutes?…” A P&G analyst answers: “…no as machines that to be broken I mean there is no fundamental reason for Procter to be down more than two percent today…” (CNBC 2010) Wall Street Journal 7 May 2010 “A bad day in the financial markets was made worse by an apparent trading glitch, leaving traders and inves- tors nervous and scratching their heads over how a mistake could send the Dow Jones Industrial Average into a 1,000-point tailspin.” “Traders theorized that an initial trading error triggered a piling-on effect from computerized trading programs designed to sell when the market moves lower. At the same time, pre-set orders from individual traders and investors to sell on declines during market downturns were likely triggered.” “The move highlighted how fragile U.S. markets have become and how the various fragmented markets have deficiencies in the way they buffer volatility.” (Lauricella and McKay 2010) Marketwatch.com 11 May 2010 “Two top financial regulators said Tuesday they aren’t sure yet what caused the stock market’s dizzying May 6 plunge and partial recovery, but they don’t believe any one event created it At issue is the Dow Jones Industrial Average drop of nearly 1,000 points last Thursday—a fall of roughly $1 trillion in market value—much of it in a matter of minutes, before recovering to a 348-point loss for the session Both Mary Schapiro, Securities and Exchange Commission chairwoman, and Gary Gensler, Commodity Futures Trading Commission chairman, refuted speculation that a trader might have made a so-called "fat finger" error that contributed to the stock market plunge.” (Orol 2010) Financial Times 14 May 2010 “Just over a week later the cause of the “flash crash”, where in the space of those 20 min stocks plunged and rebounded, still remains a mystery. Talk, however, circulates that an algorithm, or “algo” computer program that dominates trading these days, may have exploited an already nervous market by sinking the shares and then buying them back at much cheaper prices.” (Mackenzie 2010) Source References and the Authors Waddell & Reed. But in addition to that trading error, units did have a data-driven explanation about the market a number of possible reasons for the crash has since turmoil. On May 6, traders were also stunned by the sudden come to light. One of those supposed causes was high- sharp moves (Lauricella and McKay 2010). frequency trading, according to a report from the Secu- Risk and uncertainty are entirely different concepts. rities and Exchange Commission that year. Uncertainty leaves risk management techniques ineffec - tive. For financial markets, uncertainty also fuels rumors As another comment made in the fourth anniversary of the and home-made stories about market events, as the markets flash crash has been raised an idea that the causes are still experienced during and after the May 6 market crash. The not fully agreed (Krantz 2014): best strategy in these cases is getting rid of uncertainty. In Even four years after the crash that wiped out $1 tril- this regard, financial supervisors should have an automated lion in wealth in the blink of an eye, investors and and real-time data collection capacity with advanced data academics still haven’t agreed on what caused one of analytics tools as well as prudential supervisory disclosure the most vicious and inexplicable short circuiting of capacity to keep the markets running without uncertainty. market to occur. On May 6, the market participants were in a panic situa- tion, and neither market participants nor market surveillance 322 S. Zeranski, I. E. Sancak fully automated data collection system (CFTC and SEC Supervisory technological capacity of the U.S. financial authorities 2010). The May 6 market crash also an indicative market event On the other hand, there was no evidence that both agen- cies deployed advanced data analytics technologies such as in terms of the technological capacity of the U.S. financial authorities in 2010. There were and still are two main regu- artificial intelligence, machine learning, natural language processing. The above picture was as of May 2010. How- latory and supervisory agencies for the U. S. capital markets: The Securities and Exchange Commission, the SEC, and ever, Broeders and Prenio (2018) mentions data analytics tools that the SEC either uses or projects to use as of 2018. the Commodity Futures Trading Commission, the CFTC. Two agencies established a Joint CFTC-SEC Advisory Com- In response to the situation realized after the May 6 flash crash, the SEC announced a project that may eliminate leg- mittee on Emerging Regulatory Issues. The Committee’s establishment was one of twenty recommendations included acy systems, let the agency collect real-time basis data, and deploy advanced data analytics. We will handle the policy in the agencies’ joint harmonization report issued in October 2009 (CFTC and SEC 2010). The joint committee published response of the U.S. financial supervisors in the following chapters. However, the May 6 case clearly indicated that the the “Preliminary Findings Regarding the Market Events of May 6, 2010,” on May 18, 2010, twelve days after the case. U.S. financial supervisors, namely the SEC and the CFTC did not have good enough SupTech capacity at the time, even The report’s content draws a general picture about the ini- tial findings of the May 6 market crash. The same report though the U.S. is one of the leading technology innovating countries in the world. also gives the framework of the U.S. financial markets as well as the supervisory technology at the time. The follow- Since the October 1987 crash, the market structure has evolved as technological advancements have enabled par- ing statement with the CFTC and the SEC (2010) report is a summary of the SupTech capacity of the U.S. financial ticipants to trade using algorithms with little or no human intervention (Kirilenko et al. 2018). The requirements of supervisors at the time: supervisory technology for the U.S. markets have been sig- It is important to emphasize that the review of the naled since the 1980s; however, the pace of the technol- events of May 6 is in its preliminary stages and is ogy adoption has been significantly different between the ongoing. The reconstruction of even a few hours of markets and their supervisors. We observe here another fact trading during an extremely active trading day in mar- that it is not about having available technology; it is about kets as broad and complex as ours— involving thou- organizing, designing, and having a well-functioning SupT- sands of products, millions of trades and hundreds of ech and supervisory system at large. On the other hand, as millions of data points—is an enormous undertaking. we pointed out in different sections, having a full-fledged Although trading now occurs in microseconds, the SupTech system is not entirely tied to financial supervisors. framework and processes for creating, formatting, and It requires additional funds and political support as well as collecting data across various types of market partici- leadership. pants, products and trading venues is neither stand- Since the U.S. capital markets have more than one super- ardized nor fully automated. Once collected, this data visor, authorities shared the workload of analyzing the data must be carefully validated and analyzed. Such further considering their responsibility areas. In this regard, for data and analysis may substantially alter the prelimi- example, the SEC has sourced and analyzed price, time, nary findings presented in this report. The staffs of and volume data on over 19 billion shares executed on May the Commissions therefore expect to supplement this 6, and quote data representing the best bid and best offer report with further additional findings and analyses. for over 7,800 securities, for each exchange, for each mil- In summary, the U.S. financial authorities, the SEC and lisecond during the trading day, and the CFTC has analyzed transaction and order book data on stock index futures, the CFTC, in 2010, did not have. including the E-Mini S&P 500 futures contract (CFTC and • SEC 2010). Data collection, consolidation, and data analyt- real-time data collection capacity, • cross-market surveillance capacity, ics from two different channels by two different supervisors • in the same jurisdiction for the same case are not effective consolidated transaction data collection capacity, • consolidated order tracking system, supervisory strategies in the FinTech world. • After the May 6 case, the SEC emphasized the impor- standardized data, tance of having a consolidated order tracking system or consolidated audit trail system. If adopted, this rule pro- posal should result in a continuous reporting mechanism for 5 –6 market participants to capture the data needed for effective One million microseconds are equal to one second (1 µs = 10 s). Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 323 Table 3 Overall quality check of the dimensions of the U.S. financial to change this picture in a free market economy. Second, supervisory system as of May 2010 the supervisory structure is fragmented, and this also poses a significant risk for the markets. The May 6 case forced to Feature Status bring the CFTC and the SEC together on a project basis. Organizational Structure Fragmented However, later, the flash events in the U.S. Treasury markets Supervisory Model Not for the of 15 October 2014 led to a bigger coordination require- FinTech ment: the U.S. Department of the Treasury, the FED, the World Federal Reserve Bank of New York, the SEC and the CFTC. Real-Time Data Collection NA Restructuring economic and financial agencies is a national Automated Data Collection Partly economic area, and it is in the hands of the governments and Digital Identification NA politicians. It is a strategic risk management area. Thus, it is Early Warning System NA a real challenge but still possible. Regulatory and Industry Sandboxes NA Data Analytics Partly Supervisors’ technological, administrative, Prudential Supervisory Disclosure NA and policy responses Source The Authors The exchanges report the daily positions and transactions of each The CFTC and SEC’s initial report, Preliminary Findings clearing member to the CFTC and the data are transmitted elec- Regarding the Market Events of May 6, 2010, was partially tronically during the morning after the “as of” date SEC and CFTC relieving work to eliminate some rumors and cascading (2010). The CFTC also collects trade data on a daily, transaction date + 1 (“T + 1”), basis from all U.S. futures exchanges through effects of the crash; however, it was not a clear answer at Trade Capture Reports SEC and CFTC (2010) the time. The report stated the lack of. For the CFTC, all transactional data is received overnight, loaded in the CFTC’s databases, and processed by specialized software applica- real-time, tions that detect patterns of potentially abusive trades and alerts SEC standardized, and CFTC (2010) automated data collection features. cross-market surveillance (CFTC and SEC 2010). Cross- These features are the main pillars of digital financial market surveillance was not in play on May 6. Therefore, it systems in the FinTech world (Zeranski and Sancak 2020). was impossible to capture the big picture of the U.S. capi- The content of the initial report and some statements, such tal markets. Together with other lessons, the May 6 crash as “It is important to emphasize that the review of the events was an important reminder of the inter-connectedness of of May 6 is in its preliminary stages and is ongoing.” and derivatives and securities markets, particularly with respect “Much work is needed to determine all of the causes of the to index products (SEC and CFTC 2010). market disruption on May 6.”, were not so helpful to man- As specified with the paper (Zeranski and Sancak 2020), age the fear of the markets. The report’s main message was one of the most critical features of a financial supervisory not signaling a strong perception of the technology-oriented system is the real-time data collection. For example, the supervision of the markets. Instead, it was signaling that exchanges report daily to the CFCT, and the agency con- there was a considerable gap between the technology that ducts daily surveillance with them. Daily surveillance seems markets used and the supervisors’ technology. In other to be a close look at the market; however, it is not enough in words, asymmetric technology was the case. What makes the FinTech environment. the case worse, the U.S. financial markets faced an undefined Considering the main features of a digital financial system situation, and the supervisors were not in the capacity to spot set forth by the paper Zeranski and Sancak (2020), we evalu- the causes of the crash until the final report came out on Sep- ate the overall quality of the dimensions of the U.S. financial tember 30, 2010. The reform efforts and official statements supervisory system with the following table (Table 3). also signaled a long way to close the gap. Beyond the table above, the U.S. financial markets have Knowing the root causes of a market crash enables two structural weaknesses from the supervisory perspective. supervisors to respond to the drivers of the crash timely and First, the U.S. markets are fragmented. However, it is hard adequately. The May 6 case indicates that supervisors did not have SupTech tools to respond to the drivers or decide whether any additional supervisory measures should be Although fragmented markets may have many implications, in the context of the May 6 crash, Albert J. Menkveld and Bart Zhou taken on May 6 or in the aftermath. In other words, at the Yueshen’s research, The Flash Crash: A Cautionary Tale about time, neither financial supervisors nor market participants Highly Fragmented Markets, suggests that liquidity supply in severely did know what happened and why it happened exactly. Prob- fragmented markets might become vulnerable when liquidity is ably the large trader who gave momentum to the market demanded. 324 S. Zeranski, I. E. Sancak crash with its algorithmic trading also did not know what between trading venues for exchange-traded funds, happened and why it happened. equity index futures, and equity index options—instru- The May 6 case raised another question at the time: Who ments used by investors to manage their exposures in was responsible for the supervision of the May 6 market the face of broad market movements. crash? The SEC or the CFTC? Since the causes of the mar- We agree the idea stated with the CFTC and the SEC report ket crash were not known during the day, the responsible (2010) that a uniform circuit breaker rule, which would supervisor could be one of them or both of them could be. briefly pause trading across the securities markets when the Or, the worst-case scenario; none of the agencies assumes price of a security has rapidly declined over a short time, responsibility to act immediately. This question also points should make a recurrence of a severe market disruption, like out another fact: The organizational model of the U.S. finan- the one that occurred on May 6, much less likely. However, cial markets per se a source of risk in the sense that there circuit breakers neither inform financial supervisors nor mar - was an ambiguity about the responsible supervisory author- ket participants about exactly what happened at the market. ity to respond to the market crash. The responsible super- Moreover, circuit breakers may not function in every case. visor might be the CFTC or the SEC. Since there was no Designing a well-functioning circuit breaker system is a solid information about the causes of the flash crash, it was significant step to relieve extraordinary market movements. also unclear that which of them would act against the market However, it does not give a picture of abusive market trans- crash. Thus, the supervisory model of the U.S. financial sys- actions. If we assume that a circuit breaker system works tem has been improper for the FinTech world. The country’s very well, but the supervisors do not have real-time data model has been under discussion after the global financial collection capacity, in this case, supervisors still may not crisis of 2008. Today, the fast-developing FinTech sector know what happened at the market. As stated within the sec- addresses the need of reform requirements again. ond report (2010), market participants might interpret a trig- It can be said that coordination is the answer. However, gered circuit breaker with their own story and might attribute coordination between two different independent supervisory greater importance to the circuit breakers or market pauses. organizations is both an intricate issue and time-consuming Thus, we infer that particularly for fragmented and relatively in practice. In the FinTech world, market crashes require bigger financial markets, precisely for the U.S. capital mar - prompt response and reaction. There is no time to develop kets, a centralized responsible authority should inform the new formal working groups from different organizations for markets about the nature of the extraordinary market events. taking prompt actions during market crashes like the May 6 We name this kind of announcements or disclosure policy crash. Thus, cumbersome, and difficult-to-update systems, as “prudential supervisory disclosure”. As we explain it in like the U.S. financial supervisory system, are relatively a separate section, an important feature of disclosure policy riskier than the easy-to-update financial systems. is that announcements are not discretionary. Thus, market Unless financial supervisors have multi-market and multi- participants know that there will be an announcement, and asset supervision capacity, having successful supervisory they will know what is happening exactly. infrastructure at one part of the markets is not enough for To sum up, whatever was the root cause of the flash protecting market integrity. The U.S. financial supervisors crash, market participants and the public should have been have been aware of this fact and their report (2010) put it informed about that. Unless an authority, which has a capac- in this way: ity to capture the picture of the market, announces the facts An important lesson from the events of May 6 is about unusual market events, market participants and media the need to better understand cross-market linkages will produce their stories. In this regard, financial super - visors should have a SupTech capacity such that they can capture all market activities, namely orders and transactions in a real-time basis, and in certain situations, public disclo- We assume that none of the traders has an intention to cause the sure should be mandatory but not discretionary for financial market crash at the time. Since no one, including financial supervi- supervisors so as not to let media and market participants sors, has the capacity to see all the data at all markets, we assume that a single trader also cannot know the causes of the May 6 crash during use their news production vision about probable causes of the day. a market event. In fact, it was turned out that both the equity and the derivatives On the other hand, both the SEC and the CFTC took les- markets experienced severe declines and disorders on May 6. Thus, sons from the May 6 case and started new projects against the May 6 case was about both the securities market and the deriv- similar crises. atives market. That means the case was both in the SEC and in the CFTC’s areas of responsibility. In order to increase the timeliness and efficiency of There is a misconception that developed countries have always account identification, the CFTC was considering possible advantage in terms of technological reforms. In fact, this is not the rules to enhance the CFTC’s surveillance capabilities by case for every developed country. Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 325 deploying automation of the statement of reporting traders The CAT will track orders throughout their life cycle in the large trader reporting system and obtaining account and identify the broker-dealers handling them, thus allow- ownership and control information in the exchange trade ing regulators to more efficiently track activity in Eligible registers (CFTC and SEC 2010). Securities throughout the U.S. markets (FINRA 2020). On May 20, the SEC’s chairperson stated that (Schapiro, Through the CAT, regulators in the U.S. expect to have more Examining the Causes and Lessons of the May 6th Market timely access to a comprehensive set of trading data, ena- Plunge, 20 May 2010); bling authorities to more efficiently and effectively conduct research, reconstruct market events, monitor market behav- During a 20-min period during the afternoon of May ior, and identify and investigate misconduct (SEC 2019). 6, the U.S. financial markets failed to live up to their The SEC estimates that the system will cost 2.4 billion USD essential price discovery function. That period of initially and then 1.7 billion USD a year to run (Bullock and gyrating prices directly harmed those investors who Stafford 2019). The CAT project works are still ongoing, and traded based on flawed price discovery signals, and it the project has not been in play yet. undermined the confidence of investors in the integrity On the way of having a well-function SupTech system, of the markets. We are committed to taking all neces- the SEC has outsourced the Market Information Data Ana- sary steps to identify causes and contributing factors lytics System (MIDAS). The history of MIDAS began with and are already working to reduce the likelihood of a the need to more efficiently collect and analyze order book recurrence of that day. data for equities and futures (SEC 2013). According to the Some of the SEC’s proposals were about the market struc- SEC, MIDAS has many applications at the SEC, and it can ture of the U.S. capital markets. Since the market structure help the agency monitor and understand mini-flash crashes, is not the main theme of this paper, we focus on other pro- reconstruct market events, and develop a better understand- posals, mainly ones about the supervisory capacity of the ing of long-term trends. However, there are some concerns agency. about the success of the system (Podkul 2020). One of the critical steps on the way of development of The agencies’ preliminary report stated that “Although a well-functioning SupTech system has been the Consoli- the coordinated circuit breakers between futures and equities dated Audit Trail (CAT) project. In this regard, the SEC’s were not triggered, the events of May 6 reinforce the impor- chairperson stated that (Schapiro, Examining the Causes and tance of having communication links between futures and Lessons of the May 6th Market Plunge 2010); equity markets so that there is meaningful and appropriate coordination of trading pauses and halts.” One of the challenges we face in recreating the events The SEC and the CFTC’s circuit breaker project might of May 6 is the reality that the technologies used for be a good fit for the prudential supervisory disclosure sys- market oversight and surveillance have not kept pace tem. As stated by the agencies within their September 2010 with the technology and trading patterns of the rapidly report, pausing a market might be an effective way of pro- evolving and expanding securities markets.” viding a window for market participants to reassess their “Today’s fast, electronic, and interconnected markets strategies, for algorithms to reset their parameters, and for demand a robust consolidated audit trail and execution an orderly market to be re-established. In this regard, the tracking system. CME’s Stop Logic Functionality helped prevent a possibly The SEC staff started in 2009 to work, in consultation bigger and more detrimental market crash by triggering a with SROs and others, on a rule proposal that would require halt in E-Mini trading. the SROs to jointly develop, implement and maintain a con- On May 31, 2012, the SEC approved a “Limit Up-Limit solidated order tracking system, or consolidated audit trail Down” mechanism to address market volatility by prevent- (Schapiro, Examining the Causes and Lessons of the May ing trades in listed equity securities when triggered by large, 6th Market Plunge 2010). The expectations with the CAT sudden price moves in an individual stock (SEC 2012). In project were mainly to increase the ability to access in real- July 2012, the SEC also announced that the securities and time the majority of the data needed to reconstruct the type futures exchanges have procedures for coordinated cross- of the May 6 market disruption, to enhance the ability to market trading halts if a severe market price decline reaches detect and monitor aberrant and illegal activity across mul- levels that may exhaust market liquidity (SEC 2012). These tiple markets. market-wide circuit breakers may halt trading temporarily or, under extreme circumstances, close the markets before the normal close of the trading session. During the crashes, some trades might be carried out 10 with erroneous prices and later might be broken or canceled. For a summary of the SEC’s proposals against the May 6 type However, market participants cannot exactly know which of market crashes, please see Schapiro (2010). 326 S. Zeranski, I. E. Sancak their trades will be canceled, and this uncertainty may cause a contract, as the uninformed party, runs the adverse selec- further trading problems and produce additional risks. For tion risk. Therefore, information disclosure is a fundamental example, market participants might not provide liquidity in ingredient of contracts and transactions. Public disclosure such a case. As seen this an important lesson taken from the has a similar function. In a public disclosure case, one of May 6 case, to provide market participants more certainty the parties informs the unknown people, makes announce- as to which trades will be broken and allow them to better ments to the public to maintain a level playing field. There manage their risks, the SEC staff worked with the exchanges are specially designated disclosure platforms to operate pub- and FINRA to clarify the process for breaking erroneous lic disclosure activities in financial markets. For example, trades using more objective standards (SEC and CFTC EDGAR is the web-based public disclosure platform for the 2010). By using real-time data infrastructure, it can be much U.S. capital markets. An Internet-based platform enables faster to spot erroneous data and manage the broken and all parties to reach the publicized information and help col- erroneous trades. lect data. Public disclosure is one of the main features of On the other hand, in January 2020, the SEC announced capital markets, particularly stock markets—public state- that the agency would modernize the national market sys- ments and filings flood markets with other market informa- tem. The SEC’s Chairman stated that “The Commission has tion. For example, the EDGAR’s system processes about received extensive public input on issues relating to equity 3,000 filings per day, serves up 3,000 terabytes of data to market structure and access to market data, as well as sug- the public annually and accommodates 40,000 new filers per gestions for how that structure should be updated to ensure year on average (SEC 2020a, b). Therefore, data visualiza- that our markets continue to best serve the interests of inves- tion, AI, NLP, and other technology tools are now required tors. Today’s proposed order is designed to address issues to benefit from available data and information. regarding the dissemination of market data that affect the Capital markets produce in every business day, on the one efficiency and fairness of our markets. In particular, we side, massive amounts of publicized information that are welcome public input on the specific proposed governance available at the public disclosure platforms, like EDGAR, provisions.” (SEC 2020a, b). This academic paper might be on the other side, generate order and transaction data in mar- a supportive work to help increase the efficiency and fair - kets. Data vendors sell these data to their clients. Therefore, ness of the U.S. markets. However, PSD is also conducive to not everybody has all the data outreach capacity. On the other markets, particularly complex markets with multiple other hand, many financial institutions submit data to the trading venues. In this regard, the European Commission supervisors under supervisory disclosure requirements. In should also think about PSD implementation, considering this world, only several organizations can legally see all data the member states’ markets. in the financial sector. Those are financial supervisors. Financial supervisors have macro-prudential responsibili- ties. “Prudential” is literally in the meaning of “involving or Prudential supervisory disclosure showing care and forethought, especially in business” (Lex- for the digital financial world ico 2020). “Prudence” is another word for caution involving forethought, and prudential policies relate to actions that Terminology: “Prudential” , “Supervisory Disclosure” , promote sound practices and limit risk-taking (European and “Prudential supervisory disclosure” Central Bank 2017). Prudential requirements aim at mak- ing the financial sector and economy sounder and more sta- Information disclosure, public disclosure, and supervisory ble. For example, the EU rules on prudential requirements disclosure are well-known concepts in finance and the finan- mainly concern the amount of capital and liquidity of banks cial sector. Information disclosure helps contractual parties (European Commission June 2020). For the banking sector, to know about all relevant information and facts of a trans- the goal of these rules is to strengthen the EU banking sec- action or a contract. For example, a bank informs its clients tor’s resilience so that it can better absorb economic shocks when the clients would like to buy products. If one party while ensuring that banks continue to finance economic does not inform the other party fully and causes asymmetric activity and growth (European Commission 2020). information between parties, multiple risks might arise. For example, if a party does not have the full information about EDGAR is the Electronic Data Gathering, Analysis, and Retrieval As stated by the SEC and CFTC’s report (2010), on September 10, system used at the SE (SEC 2020a, b). Containing millions of com- the SEC approved the new trade break procedures, which like the cir- pany and individual filings, EDGAR benefits investors, corporations, cuit breaker program, is in effect on a pilot basis through December and the U.S. economy overall by increasing the efficiency, transpar - 10, 2010. ency, and fairness of the securities markets (SEC 2020a, b). Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 327 Mishkin (2000) broadly defines prudential supervision world, the EU should consider the PSD model for the Union- as “government regulation and monitoring of the banking wide risk management mechanism against market-driven system to ensure its safety and soundness”. The forms of risks in addition to the investment firms-based prudential prudential supervision might be. supervision. The European Banking Authority (EBA) could play a restrictions on asset holdings and activities, leading role in networking to help initiate a well-function- separation of banking and other financial industries like ing SupTech system across the Single Market (EBA 2020). securities, insurance, or real estate, The EBA considers that the European Forum for Innovation restrictions on competition, Facilitators (EFIF) provides a good means for supervisors to capital requirements, share experiences on a cross-sectoral basis, aiding the iden- risk-based deposit insurance premiums, tification of innovation trends, regulatory and supervisory disclosure requirements, issues that require a cross-sectoral position and to moni- bank chartering, tor interconnectedness on a multi-disciplinary basis (EBA bank examination, 2020). supervisory versus regulatory approach (Mishkin 2000). There are supporting views of the banking sector of the EU that FinTech activities give rise to not only operational Mishkin’s approach to prudential supervision is mainly risks but also financial risks, especially of a systemic nature related to the banking sector. However, his approach does (EBA 2018). And, a potential ‘FinTech bubble’ was raised as not consider the speed and big data factors that markets face another issue of concern with the risk of reducing the effec- today. Two decades later, we have today completely different tiveness of monetary policy noted as another potential threat financial technology. Hence, prudential supervision can be (EBA 2018). Thus, the PSD model should be considered in used for many more areas and may have a broader mission response to FinTech risks. if we enter other avenues of the financial sector. Prudential supervision can be classified into micro- and Wall (2016) addresses advanced analytics and the impor- macro-prudential supervision. The prefix “macro” indicates tance of data by stating that the availability of more granular that the policies or actions relate to the whole or significant information, combined with new tools to analyze them, may parts of the financial system rather than individual financial provide supervisors with a variety of opportunities to evalu- institutions (European Central Bank 2017). Supervisory or ate the risk of financial systems better. The development of regulatory policies for individual financial institutions, by machine learning using deep learning techniques raises the contrast, are known as micro-prudential policies (European possibility that supervisors will be able to use granular data Central Bank 2017). While macro-prudential policies con- to better understand the risks in the financial system (Wall sider the soundness of the whole financial system, soundness 2016). and informational efficiency of markets should also be in this Sound prudential supervision policies should take into scope. However, today’s macro-prudential policies do not account the potential for investment firms and their clients focus on financial market-driven informational imbalances, to engage in excessive risk-taking and the different degrees particularly FinTech environment data production, usage, of risk assumed and posed by investment firms (European and the speed factors at the electronic markets. Parliament 2018). With the PSD model, we also raise the In other words, the FinTech world brings new responsi- idea that not only financial institutions but also markets bilities to financial regulators and supervisors, as financial should also be the realm of prudential supervision for the institutions have the responsibility of Know-Your-Customer, financial regulators and supervisors. According to a report financial supervisors should also have “Know-Your-Mar - of the European Parliament (2018), differences in the appli- kets”, “Know-Your-Technology”, “Know-Your-Data”, and cation of the existing framework in different member states “Inform-Your-Markets” responsibilities. The truly trans- of the EU threaten the level playing field for investment formative potential of regulatory technology addresses a firms within the Union, hampering investors’ access to new Know-Your-Customer mindset transformation into a Know- opportunities and better ways of managing their risks. This Your-Data approach (Arner et al. 2016). FinTech crises are also holds for technology and the market data or prudential a new source of systemic risks and the market data are sys- supervisory disclosure. temically important in the FinTech world. In this regard, all Since there might be differences in prudential supervisory market data should be considered as a new area for macro- capacities among the EU member states, creating a mecha- prudential supervision. nism of cooperation and exchange of information among Markets produce massive amounts of data every day, the financial authorities to ensure harmonized prudential and data analytics tools should be deployed at supervisory supervision of investment firms across the Union seems to agencies. Without data analytics tools and a well-designed be essential (European Parliament 2018). In the FinTech SupTech system, big data might be a black hole for financial 328 S. Zeranski, I. E. Sancak supervisors. By knowing the market data and conducting but supervisors do not have specific disclosure require- data analytics, financial supervisors can utilize big data ments regarding market events. For example, Turkey’s and even stay ahead of markets. The PSD model requires main supervisor for capital markets, the CMB, is not under a well-designed SupTech system. The PSD model has both a responsibility to inform the markets publicly under a writ- a systemic risk management capacity and implications for ten disclosure policy. There are some obscure and general transparency requirements. requirements, but there is not a specific requirement and a The growing FinTech world addresses new business mod- policy document or a guide to inform related parties at the els for private sector firms and new supervision models for Turkish capital market. Therefore, market participants do not financial supervisors. In this regard, based on the prudential exactly know when the CMB will inform them or whether supervisory disclosure (PSD), we introduce the PSD model. the CMB will make any announcement or not. In the FinTech world, supervisors should react promptly The scope, design, and functions of prudential to market events. Any delayed reaction might cause inevita- supervisory disclosure ble losses and market crashes. The May 6 market crash is the case indicating the timing concern in the FinTech world. On Supervisory disclosure or public disclosure are well-known May 6, when markets were already under stress, a sell algo- concepts in the financial industry or academic world. These rithm chosen by a large trader to only target trading vol- are regulatory requirements for market participants but not ume, and neither price nor time, executed the sell program for supervisors or financial authorities. extremely rapidly in just 20 min (SEC and CFTC 2010). Financial services providers are required to disclose some During the May 6 market crash in the U.S., there were many information either to supervisors or to the public. On the unknowns and fears of unknowns about probable roots of other hand, public disclosure is one of the main responsi- the crash, and supervisory agencies were under stress to act bilities of publicly held companies. There are strict rules against the drivers of the crash. They could not reveal useful for publicly held companies to reveal proper information information to the public timely. Six days after the crash, timely for related parties; shareholders, investors, and oth- the CFTC and the SEC published a report about the market ers. In all modern financial markets, public disclosure is one events. However, the report did not exactly answer the ques- of the main features and regulated areas. For example, the tions. Twenty days later, on May 26, 2010, the chairperson Transparency Directive (2004/1009/EC) requires issuers of of the SEC made a statement and pointed out that the SEC, securities traded on regulated markets within the EU to make at the time, could not track data across multiple markets, their activities transparent, by regularly publicizing certain products, and participants in a real-time basis. Later, to fill information (European Commission 2020). In this regard, the gap, the SEC introduced a Consolidated Audit Trail pro- the information to be publicized includes (European Com- ject or CAT in short. The following statement of the chair- mission 2020): person of the SEC on May 26, 2010, inter alia, addresses the prudential supervisory disclosure requirement: yearly and half-yearly financial reports "If adopted, this consolidated audit trail would, for the major changes in the holding of voting rights first time ever, allow the SEC and other market regulators ad hoc inside information, which could affect the price to track trade data across multiple markets, products, and of securities. participants in real-time," "It would allow us to rapidly reconstruct trading activity On the other hand, as part of the supervisory disclosure and quickly analyze both suspicious trading behavior and requirements in the EU, according to the Directive 2013/36/ unusual market events." (SEC 2010a, b). EU (Capital Requirements Directive—CRD IV), all EU About the addressed May 6, 2010 crash, the SEC, together member states are required to present information regard- with the CFTC, revealed the full report approximately five ing the laws, regulations, administrative rules and general months later, which was an extremely late action in the Fin- guidance in the field of prudential regulation and supervision Tech world. On May 6, 2010, the U.S. markets suffered not (BaFin&Deutsche Bundesbank 2020). only from an improper algorithm fueled the market crash Publicly held companies and financial services provid- ers, even real person investors in some cases, are under the requirement of public disclosure or supervisory disclosure, To analyze the announcement policies of financial supervisors, we can check the official web sites since supervisors use mainly their web sites to make announcements. Periodic bulletins, press releases, and annual reports are not in the focus of these discussions. The term “Prudential Supervisory Disclosure”, or PSD, was first coined by Zeranski and Sancak (2020). This section mostly depends Algorithmic trading and high-frequency trading are in the domain on their papers. of FinTech. Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 329 but also rumors and detrimental speculations about the driv- ers of the market crash at the time. Prudential supervisory disclosure could save the market from huge material losses and loss of confidence at the time. Prudenal Supervisory It might be less detrimental if some announcements were Supervisory Disclosure by made immediately, and some supervisory actions were taken Disclosure by market during the day. However, many supervisory agencies are not Supervisory parcipants to under the obligation of revealing timely information about Agencies and Supervisory the roots of similar market crashes or events to the public to SROs to the Agencies or Public SROs or the calm down market participants, and markets at large. They Public do this only within their discretion. As in the May 6 case, some supervisors had not been in that capacity, too. There- fore, we see this situation as a new risk for financial markets in our high-speed FinTech world. Public disclosure is mainly a pillar of market discipline and transparency issues. The BCBS touches on supervisory disclosure from the banking sector perspective in the report, Fig. 2 Prudential supervisory disclosure Enhancing Bank Transparency: Public Disclosure and Supervisory Information that Promote Safety and Sound- the markets have all the required information or will have ness in Banking Systems (BIS 1998): “Market discipline, however, can only work if market it timely. The following concerns address the requirement of pru- participants have access to timely and reliable information which enables them to assess a bank’s activities and the dential supervisory disclosure: risks inherent in those activities. Improved public disclosure strengthens market participants’ ability to encourage safe (1) Timing Concern: FinTech has the potential to bring a more radical change within the financial industry and sound banking practices.” We believe that there are multiple drivers for prudential and become a core constituent of its infrastructure and processes, hence boosting the speed and the agility of supervisory disclosure. For example, the global financial cri- sis (2007–2009) has called into question the role of finan- financial services (Kashyap and Weber 2018). In the FinTech world, transactions take place in microsec- cial policy in general, especially in banking, revealing major shortcomings in market discipline, regulation, and supervi- onds. Any late response to the markets might cause severe crashes, FinTech crises and financial crises. sion (The World Bank 2020). Additionally, in the FinTech world, supervisors may have more technological opportuni- (2) Scope Concern: Publicly held companies and FSPs have limited scope, not a market-wide scope. In some ties and potentially strong tools available to carry out their duties. The FinTech world also brings new responsibilities of cases, there might be material information that only supervisors have with their vast information outreach more active market surveillance and more promptly response to the market crashes and abusive market transactions or capacity from multiple data sources. In many regu- latory frameworks, banks, FSPs at large, transmit to news. Moreover, supervisors may have a bigger capacity to contribute to financial stability. From the market quality supervisory authorities, based on a relationship covered by professional secrecy laws and rules, a larger amount perspective, supervisors can contribute more to the informa- tional efficiency of the markets. of accounting data and other information than they are legally required to make public (e.g., annual reports) The structuring a prudential supervisory model is not a complicated work, but it requires a new supervision perspec- or that they publish voluntarily (e.g., in the press) and supervisory authorities can use this important stock of tive. In the private sector, financial services providers have been changing their business models in the FinTech world. information not only to perform the tasks entrusted to them by law but also to enrich the information avail- Their supervisors should also update their business models, or, their supervision perspectives accordingly. able to the public (BIS 1998). As stated with the same BIS report, confidentiality will not be breached if the The following figure shows an interaction between two parties in terms of the prudential supervisory model (Fig. 2). information is released in aggregate forms. In this regard, each supervisory authority should have a predetermined and written prudential supervisory disclo- sure policy. Under this policy, market participants know that 16 –6 One million microseconds are equal to one second (1 µs = 10 s). 330 S. Zeranski, I. E. Sancak (3) Technology Concern: Supervisors potentially have more technological tools and solutions to carry out their duties. SupTech gives supervisors both more Financial Supervisors: Prudenal technological tools and new responsibilities. Comput- Supervisory Disclosure for Their Jurisdicons/Sectors ers can exceed the abilities of human experts in some cases (BaFin 2018). 4) Transparency Concern: Market participants should know under a regulation that supervisors are under the obliga- SROs: Prudenal Suprevisory Disclosure for Their Markets tion of revealing all market-sensitive information, and supervisors will reveal all relevant information to the public. Transparency concern also addresses the infor- Financial Services Providers, Publicly- mational efficiency. traded Companies, Large Shareholders: 5) Mandate Concern: Each financial supervisory authority Supervisory Disclousure/Informaon Disclosure/Public Disclosure has designated mandates and is responsible for success- fully fulfilling the mandates. SupTech applications can turn risk and compliance monitoring from a backward- looking into a predictive and proactive process (Broed- ers and Prenio 2018). Solutions that use advanced data Fig. 3 Laddering PSD analytics and technologies could lead to more timely, dynamic, and even predictive supervision, which enables supervisors to extract knowledge from data that would enable a level playing field for all market participants, and eliminate technology-related asymmetric information risks. be otherwise inaccessible (Dias 2017). Having higher data collection capacity and data analytics tools forces The responsibility of prudential supervisory disclosure might be delegated to SROs when the prudential information supervisors to deliver more useful products and services in a timely fashion. is only in the hand of an SRO. The following figure indicates a possible delegation model (Fig. 3). (6) Accountability Concern: Supervisors should be accountable for their poor disclosure policies to pro- Under the current setup, some supervisors already share their supervisory responsibilities with SROs, such as FINRA tect market integrity and financial consumers as well as financial stability. and exchanges in the U.S. Therefore, a centralized SRO might be assigned as the PSD agency in this regard. Twenty years ago, the regulatory sandbox idea was not A partly similar rule to PSD has already been in prac- tice at the U.S. markets. The Rule 603(b) of Regulation a vision in the supervisory landscape. However, it is well accepted today. Prudential supervisory disclosure is only National Market System (NMS) requires equity exchanges and FINRA to act jointly to disseminate consolidated infor- another concept that we may face soon to handle FinTech related issues properly. mation, including a National Best Bid and Offer (NBBO), on quotations for and transactions in NMS stocks (SEC and To sum up, a SupTech system enables supervisors to collect much better information timely. And, by hav- CFTC 2010). The rule is mainly for fair trading practices. In this regard, the consolidated information is disseminated ing market-wide information, supervisors should reveal information as part of a prudential supervisory disclosure through securities information processors that collect, pro- cess and prepare to publish such information, including the policy to calm down markets, especially in stressful times, increase informational efficiency, cope with market-wide price, size, and symbol of quotations and executions (SEC and CFTC 2010). rumors and increase confidence in the markets. We assume that the initial concern about the PSD model The SEC rules require that the exchanges and FINRA provide timely and accurate data to the Consolidated Tape might be the operational responsibility. It is an acceptable concept that regulators and supervisors should not act as if System (CTS) and Consolidated Quotation System (CQS) to inform all participants of the trading and quoting activi- they are a market actor. They might not be a market actor, but they are also not outside of the markets. Regulators and ties occurring in the market place (SEC and CFTC 2010). According to the SEC and the CFTC’s report (September supervisors and central banks are not entirely outside of the daily market operations. They interfere with markets for the 2010), at the time, there was considerable attention in the public media regarding the data delays, and the staff agreed sake of market integrity and financial stability. On the other hand, staying technological-neutral, with the PSD model, that this was an important topic that should be addressed. PSD is a broader concept and requires informed market par- financial supervisors do not affect the market directions, but contribute to the informational efficiency of the markets, ticipants not only for orders and transactions at one exchange Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 331 or trading center but also for capital market-wide material not obliged to inform financial consumers with all available information that only the supervisory agencies have by their information about banks and banking sector. mandate. After the global financial crisis, bank supervision became Prudential supervisory disclosure is the name of public stricter and more complex, and supervisory capacity did not disclosure for supervisors in the FinTech world. It is a tech- improve proportionally to match the greater complexity of nical requirement with SupTech today. A PSD model might bank regulations (The World Bank 2020). On top of that, the be the idea of flying cars for today, but it seems a reality for SupTech capacity of many financial supervisors could not the future. A statistic says that ninety percent of the data in catch up with the FinTech developments. the world was created in the previous two years alone (IBM As one of the drivers of the global financial crisis, the 2016). Therefore, we should not extrapolate the past too far risk was transferred in nontransparent ways owing to the into the future for technological developments. rapidly increasing trade in complex, structured financial products (The World Bank 2012). Today, some risks might Implications of prudential supervisory disclosure be within the big data, and unless data analytics tools cap- for banking and capital market sectors ture these risks and timely published by the relevant parties to the public, uncertainty might fuel some other FinTech The global financial crisis (2007–2009) has called into crises. Therefore, Know-Your-Data and PSD are crucial for question the role of financial policy in general, especially in financial stability. banking, revealing major shortcomings in market discipline, Using data on publicly traded banks in 61 countries, regulation, and supervision (The World Bank 2020). Fin- Anginer et al. (2018) examined how the institutional envi- Tech also increases the importance of totally new financial ronment affects the relationship between bank capital and policies. The pace of technology, as well as some FinTech system-wide fragility. Their research concludes that bank crashes, addresses the urgency of supervisory reforms. The capital is associated with a reduction in the systemic risk May 6 market crash has many lessons for both capital mar- contribution of individual banks and this effect is more pro- kets and banking sector supervisors. nounced for banks located in countries with less efficient Our analyses mainly focus on the May 6 market events, public and private monitoring of financial institutions and which are about the capital market sector. We prove that the in countries with lower levels of information availability lack of a well-designed SupTech system leaves the capital (Anginer et al. 2018). markets unprotected in the FinTech world. Many economies The study of Demirgüç-Kunt et al. (2008) finds a signifi- still run the same risks today. cant and positive relationship between compliance with the On the other hand, the PSD model also has many implica- Core Principles for Effective Banking Supervision related tions for the banking sector. The decade following the global to information provision and bank soundness. Countries financial crisis was characterized by intense regulation of that require their banks to regularly and accurately report banking sectors worldwide, especially in advanced countries their financial data to regulators and market participants (The World Bank 2020). A decade after the global financial have more highly-rated banks, as timely disclosure of high- crisis, intense regulation seems to be not enough to keep the quality information strengthens monitoring by regulators financial sector safe and sound, since technological develop- and markets alike (Demirgüç-Kunt et al. 2008). Their results ments have been disrupting the sector and transcending the suggest that countries aiming to upgrade banking regulation regulatory issues. and supervision should consider giving priority to informa- The PSD model may be more conducive to the banking tion provision over other elements of the core principles. sector since public disclosure rules and regulations are not similar to the publicly traded companies, which are in the Challenges for financial supervisors realm of the capital market sector. In other words, banks have more regulatory rules to submit information to supervi- In today’s world, in addition to the low pace of digital sors than to financial consumers about their capital require- transformation of supervision with SupTech, one of the ments, operations, organizations, and financial soundness. main challenges is the fragmentation of financial supervi- Therefore, n fi ancial supervisors collect colossal information sion. Supervision is divided among the FED, the FDIC, the from the banking sector, much of them are not available OCC, the SEC, FINRA, the CFTC, and state regulators in for financial sector participants. Moreover, after the global the U.S., centralized bodies such as the ECB, the SSM, the financial crisis, bank regulations became more complex, ESMA, the EIOPA, and the EBA share a stage with national potentially reducing transparency, increasing regulatory arbitrage, and taxing supervisory resources and capacity (The World Bank 2020). However, financial supervisors are We still reserve the privacy issues of banks. A sample of 39 countries. 332 S. Zeranski, I. E. Sancak competent authorities in the EU, and regulation is developed in the next fiscal year. Assuming success in the second fiscal at a national level, and regional coordination is limited in year, two years delay without technology investment makes Asia (Frisell et al. 2018). Regulatory and institutional frame- supervisory agencies old-fashioned in the FinTech world. works will need to be revised in light of new and evolving This picture leaves supervisory agencies unarmed against risks and industry landscapes (Frisell et al. 2018). In other the fast-growing FinTech world. words, as pointed out with a recent paper, Digitalization of What makes the case worse is that operating in a tech- Financial Supervision with Supervisory Technology, before nologically leading country does not help the U.S. financial the digital transformation, a check-up for the whole financial supervisors carry out their responsibilities successfully, but system and adjustments for the financial structure are the the technology stirs mostly financial markets and institu- prerequisites of having a modern and functional supervisory tions, namely the private sector. This legacy political struc- system. ture is per se a source of risk for financial stability. As we The regulatory and supervisory framework in the EU observe that the major financial reforms have been followed does not directly address the RegTech or SupTech para- mostly by crises or scandals, unless politicians do not get digms, and the approach taken by firms and supervisors to pressure from lobbying channels, they are not inclined to pilot and adopt RegTech and SupTech frameworks is cur- increase the budget for financial supervisors. This situation rently ad-hoc and uncoordinated (European Commission also seems a kind of vicious circle or dilemma for the stabil- 2019). This was seen as an important issue and handled by ity of the financial industry and the global economy at large. a report, Expert Group on Regulatory Obstacles to Financial Innovation (ROFIEG):30 Recommendations on Regulation, Innovation and Finance -Final Report to the European Com- Conclusion mission. The Group recommends that the EU develops and implements a comprehensive and ambitious agenda for the Financial supervisors collect vast amounts of data and infor- establishment of advanced RegTech and SupTech capabili- mation about market institutions and activities, but they are ties, in coordination with relevant authorities in and beyond not obliged to reveal information under a specific disclosure the EU and international standard setters. policy. They inform the public within their discretion but It is estimated that finance goes real-time, and periodic not under a predetermined disclosure policy. This should reporting no longer drives operations and decisions in the not be the case anymore in the FinTech world with a SupT- near future (Deloitte 2020). And, speed has always been of ech capacity. As financial institutions have the responsibil- the essence in financial markets (Ait-Sahalia and Saglam ity of Know-Your-Customer, financial supervisors should 2013). Therefore, supervisors also should equip with high- also have “Know-Your-Markets”, “Know-Your-Technology”, speed technological tools to respond to market crashes and “Know-Your-Data”, and “Inform-Your-Markets” responsi- FinTech crises properly. bilities. With the PSD model, from the market quality and FinTech players often fall outside the applicable regu- micro-structure perspectives, supervisors can contribute latory and supervisory framework both for prudential and more to the informational efficiency of the markets. customer protection supervisions, which is a challenge that The U.S. financial markets faced an unprecedented rapid regulators or supervisors with capacity constraints may be decline and recovery on May 6, 2010, known as the May ill-equipped to address (Berg et al. 2020). The U.S. finan- 6 flash crash. Roughly one trillion $ market value in less cial markets are the most complex markets from supervisory than thirty minutes vanished with the biggest one-day point perspectives. The magnitude of transactions, fragmented decline in the history of the DJIA at the time. Since the but interconnected markets, the variety of financial instru - market events took place in electronic markets, and algo- ments coupled with the fragmented and intricate design of rithmic trading (AT) and high-frequency trading (HFT), the financial regulatory and supervisory structure make the parts of FinTech, played significant roles, we handle the markets incredibly difficult to monitor, manage FinTech May 6 flash crash from the FinTech, SupTech, and financial related risks, and cope with financial frauds. supervision perspectives. Our research is unique because Even though financial supervisors are independent in we analyzed the May 6, 2010 flash crash first time from their responsibility areas, their budgets are tied to political FinTech and SupTech, or “TECHs in Finance” perspectives. decisions. In many countries, budget allocations to inde- We flashbacked the events and analyzed the responses of the pendent financial agencies take place only once a year. Since economic and financial media and two U.S. financial super - the financial sector of the U.S. is extraordinarily complex, visors, the SEC and the CFTC, to the market events. The modernization and updating works require not millions but case has many lessons and takeaways for the governments, billions of U.S. dollars. If an agency cannot get a budget economic policymakers, and regulatory and supervisory increase for technology investments or reform requirements authorities, and academic communities. Our analyses are for a fiscal year, then the agency has the chance to get it only Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 333 more conducive to the U.S. and the EU because their frag- PSD capacity, the EU Member States, the U.S., and many mented financial systems are in the urgent need of TECHs other countries run similar risks. in Finance reforms. Analyzing the May 6 flash crash, we find that the techno- Funding Open Access funding enabled and organized by Projekt logical imbalance between financial markets or institutions DEAL. and their supervisors drove the markets in uncertainty, hence in a fear and panic environment. Since the imbalance has not Compliance with ethical standards diminished yet, the same risks still exist. As a remedy, we introduce a new concept, prudential supervisory disclosure Conflict of interest The authors certify that they are not affiliated with (PSD), and a model, the PSD model, with a well-function- or involved in any organization or entity with any financial interest or ing SupTech system, to cope with the May 6 type FinTech non-financial interest in the subject matter or materials discussed in this manuscript. crises. Even though the U.S. has been one of the leading technology innovating countries in the world, the May 6 Open Access This article is licensed under a Creative Commons Attri- case indicated that the U.S. financial supervisors, namely the bution 4.0 International License, which permits use, sharing, adapta- SEC and CFTC did not have good enough SupTech capac- tion, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, ity at the time. We are convinced that it is not about having provide a link to the Creative Commons licence, and indicate if changes available technology; it is about organizing, designing, and were made. The images or other third party material in this article are having a well-functioning SupTech and supervisory system included in the article’s Creative Commons licence, unless indicated at large. Moreover, having a full-fledged SupTech system is otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not not wholly tied to financial supervisors. It requires additional permitted by statutory regulation or exceeds the permitted use, you will funds and hence political support as well as leadership. need to obtain permission directly from the copyright holder. To view a Risk management policies were developed mostly after a copy of this licence, visit http://creativ ecommons .or g/licenses/b y/4.0/. crisis comes out in the financial sector. However, we do not have such a comfortable reform approach any more against FinTech crises. Due to the nature of such crises, markets and institutions can be wiped out in hours, if not in minutes. References The May 6 market crash depleted market liquidity in twenty minutes, collapsed prices, and caused a massive panic at the Ait-Sahalia, Y., and M. Saglam. 2013. High Frequency Traders: Taking U.S. financial markets. Advantage of Speed. Cambridge: National Bureau of Economic Research. PSD is the name of public disclosure for supervisors in Anginer, D., A. Demirgüç-Kunt, and S.D. Mare. 2018. Bank capital, the FinTech world. It is a technical requirement with SupT- institutional environment and systemic stability. Journal of Finan- ech today. The PSD model helps protect market integrity by cial Stability, pp. 97–106. revealing useful information timely about market functions Arner, D.W., J. Barberis, and R.P. Buckley. 2016. The emergence of RegTech 2.0: from know your customer to know your Data. Jour- and against improper market activities or rumors. A PSD nal of Financial Transformation, pp. 79–86. model might be the concept of flying cars for today, but it BaFin, F. 2018. Big Data Meets Artificial Intelligence: Challenges and seems a reality for the future. Taking seriously a statistic Implications for the Supervision and Regulation of Financial Ser- about 2017 trends saying that ninety percent of the data in vices. Bonn: BaFin. B.D. Bundesbank. 2020. Supervisory Disclosure. Retrieved from the world was created in the previous two years alone, we do Supervisory Disclosure: https ://www.super visor y-discl osure .de/ not extrapolate the past too far into the future for technologi- super visor y-en/ cal developments. Thus, we expect the PSD model or a ver- Berg, G., M. Guadamillas, H. Natarajan, and A. Sarkar. 2020. Fintech sion of the model as the next normal of the financial sector. in Europe and Central Asia: Maximizing Benefits and Managing Risks. Washington, DC.: The World Bank Group. FinTech crises might cause bank runs and destroy the BIS. 1998. Enhancing Bank Transparency: Public Disclosure and baking sector as well as capital markets. One of the initial Supervisory Information that Promote Safety and Soundness in considerations for the May 6 market crash was about fat Banking Systems. Basle: Bank for International Settlements-Basle finger speculations for a bank. This makes the case more Committee on Banking Supervision. Borch, C. 2017. High-frequency trading, algorithmic finance, and the important; banks are vulnerable to rumors which might trig- flash crash: reflections on eventalization. Economy and Society, ger bank runs. In this regard, the PSD model is crucially pp. 350–378. important to protect banks from FinTech crises and bank Bouveret, A., P. Breuer, Y. Chen, D. Jones, and T. Sasaki. 2015. Fra- runs. We contemplate that the May 6 case could have been gilities in the U.S. Treasury Market: Lessons from the “Flash Rally” of October 15, 2014. Washington. D.C.: International more detrimentally and driven the banks into collapse under Monetary Fund. the supervisory setup at the time. And, it could have been Broeders, D., and J. Prenio. 2018. Innovative Technology in Financial less detrimental under a PSD model. Currently, without a Supervision (Suptech)-the Experience of Early Users. Basel: Bank for International Settlements-Financial Stability Institute. 334 S. Zeranski, I. E. Sancak Bullock, N., and P. Stafford. 2019. SEC Approves Vast Surveillance Kirilenko, A.A., A.S. Kyle, M. Samadi, and T. Tuzun. 2017. The flash System for Stock Market. Retrieved from Financial Times: https: // crash: high-frequency trading in an electronic market. Journal of www.ft.com/conte nt/2a428 156-ab1e-11e6-9cb3-bb820 79021 22. Finance, pp. 967–998. CFTC and SEC. (2010). Preliminary Findings Regarding the Market Kirilenko, A.A., A.S. Kyle, M. Samadi, and T. Tuzun. 2018. Automa- Events of May 6, 2010. washington, D.C.: CFTC and SEC. tion, intermediation and the flash crash. Journal of Investment CNBC. (2010). FLASH CRASH May 6, 2010 (Part 6 of 6) CNBC. Management, pp. 29–46. Retrieved from Youtube: https ://www.y outu be.com/w atc h Krantz, M. (2014). Four-year flash crash anniversary haunts markets. ?v=7UhKO s3dYk 4. Retrieved from CNBC: https ://www.cnbc.com/2014/05/05/four- CNBC. 2014. The Lasting Impact of the 2010 Flash Crash. Retrieved year-flash -crash -anniv ersar y-haunt s-marke ts.html. from CNBC: https ://www .cnbc.com/2014/05/06/the-las ti ng-im pac Krudy, E. (2010). Stocks plunge as trading glitch suspected. Retrieved t-of-the-2010-flash -crash .html. from Reuters.com: https ://www.reute rs.com/artic le/us-marke ts- Deloitte, J.A. 2020. Digital Finance Series: Finance 2025. London: stock s/stock s-plung e-as-tradi ng-glitc h-suspe cted-idUST RE634 Deloitte.1EA20 10050 6. Demirgüç-Kunt, A., E. Detragiache, and T. Tressel. 2008. Banking on Lauricella, T., and P.A. McKay. 2010. Dow Takes a Harrowing the principles: compliance with Basel core principles and bank 1,010.14-Point Trip; Biggest Point Fall, Before a Snapback; Glitch soundness. Journal of Financial Intermediation, pp. 511–542. Makes Thing Worse. Retrieved from WSJ: https ://www.wsj.com/ Dias, D. 2017. FinTech, RegTech and SupTech: What They Mean for artic les/SB100 01424 05274 87043 70704 57522 77541 31412 596. Financial Supervision. Toronto: Toronto Centre. Lexico. 2020. Definition of prudential in English. Retrieved from Lex- Easley, D., M. Lopez de Prado, and M. O’Hara. 2011. The microstruc- ico: https ://www.lexic o.com/en/defin ition /prude ntial . ture of the ‘Flash Crash’: flow toxicity, liquidity crashes and the Mackenzie, M. 2010. Vital lessons of the ‘flash crash’. Retrieved probability of informed trading. The Journal of Portfolio Manage- from ft.com/: https ://www.ft.com/conte nt/2e1a9 6ee-5f7e-11df- ment, pp. 118–128.a670-00144 feab4 9a. EBA. 2018. The EBA’s FinTech Roadmap: Concluions from the Consul- Menkveld, A. J., and B.Z. Yueshen. 2018. The flash crash: a caution- tation on the EBA’s Approach to Financial Technology (FinTech). ary tale about highly fragmented markets. Management Science, Paris: EBA. pp. 4470–4488. EBA. 2020. EBA Response: EC consultation on the Digital Finance Mishkin, F. S. 2000. Prudential Supervision: What Works and What Strategy/Action Plan. Paris: EBA. Doesn’t? NBER Conference-Working Paper 7926 (pp. 1–36). European Central Bank. 2017. A Quick Guide to Macroprudential Poli- Florida: National Bureau of Economic Research. cies. Retrieved from European Central Bank: https ://www.ecb. Orol, R.D. 2010. SEC: No evidence that one event caused market eur op a.eu/expla iners /tell-me-mor e/html/macr o pr ude ntial polic plunge. Retrieved from Marketwatch.com: https ://www .mar k e ies.en.html.twatch.com/s tory/no-e vidence-of-f at-finger -error-in-ma y-6-plung European Commission. 2020. Prudential Requirements. Retrieved from e-2010-05-11. European Commission: https ://ec.europ a.eu/info/busin ess-econo Peress, J. 2014. The Media and the Diffusion of Information şn Finan- my -eur o/banki ng-and-f inan ce/f inan cial-super visio n-and-r isk - cial Markets: evidence from Newspaper Strikes. The Journal of manag ement /manag ing-r isk s -bank s -and-finan cial-ins ti tutio ns/ Finance, pp. 2007–2043. prude ntial -requi remen ts_en. Podkul, C. 2020. When Volatility Surges, SEC’s Trade-Monitoring Sys- European Commission. 2019. Expert Group on Regulatory Obstacles tem Has Struggled. Retrieved from WSJ: https ://www.wsj.com/ to Financial Innovation (ROFIEG):30 Recommendations on Reg-articles/when-v olatility -sur ges-secs-tr ade-monit or ing-sy stem-has- ulation, Innovation and Finance -Final Report to the European strug gled-11585 79644 2?mod=searc hresu lts&page=1&pos=5. Commission—December 2019. Brussel: European Commission. Schapiro, M. L. 2010. Testimony Concerning the Severe Market Dis- European Commission. 2020. Transparency Requirements for Listed ruption on May 6, 2010. Retrieved from SEC: https ://www.sec. Companies. Retrieved from European Commission: https ://gov/news/testi mony/2010/ts051 110ml s.htm. e c. e ur o p a .e u /i n fo /b u si n e ss - e co n o my- eu r o/ c om pa ny- r e p or t in g - Schapiro, M.L. 2010. Examining the Causes and Lessons of the and-auditing/com pan y-reporting/tr anspar ency -requir ements-lis te May 6th Market Plunge. Washington, D.C.: U.S. Securities and d-compa nies_en. Exchange Commission. European Parliament. 2018. Report on the proposal for a directive SEC. 2010. SEC Proposes Consolidated Audit Trail System to Better of the European Parliament and of the Council on the pruden- Track Market Trades. Retrieved from SEC: https ://www.sec.gov/ tial supervision of investment firms and amending Directives news/press /2010/2010-86.htm. 2013/36/EU and 2014/65/EU. Retrieved from https:/ /www.europ SEC. 2012. Investor Bulletin: Measures to Address Market Volatil- ar l.eur op a.eu/: https ://www .eur op ar l.eur op a.eu/doceo /docum ity. Retrieved from SEC: https ://www.sec.gov/oiea/inves tor-alert ent/A-8-2018-0295_EN.pdf.s-bulle tins/inves tor-alert s-circu itbre akers bulle tinht m.html. FINRA. 2020. catnmsplan. Retrieved from FINRA CAT: https://www . SEC. 2013. MIDAS-Market Information Data Analytics System. catnm splan .com/. Retrieved from SEC: https://www .sec.gov/markets tructur e/midas Frisell, L., M.J. Fernandes, L. Quest, E. Rennick, S. Roy, and D. .html#.XoTYl 4hKg_E?mod=artic le_inlin e. Treeck. 2018. Supervising Tomorrow. New York: Oliver Wyman. SEC. 2019. SEC Approves Plan to Create Consolidated Audit Trail. Gomber, P., J.A. Koch, and M. Siering. 2017. Digital finance and Fin- Retrieved from U.S. Securities and Exchange Commission: https Tech: current research and future research directions. Journal of ://www.sec.gov/news/press relea se/2016-240.html. Business Economics, pp. 537–580. SEC. 2020. About EDGAR. R etrieved from SEC: https://www .sec.gov/ IBM. 2016. 10 Key Marketing Trends for 2017. New York: IBM.edgar /about . Janjigian, V. 2010. Fat Fingers Cause Panics. Retrieved from Forbes. SEC. 2020. SEC Proposes Improvements to Governance of Market com: https://www .forbes.com/sites /g reatspecu latio ns/2010/05/06/ Data Plans. Retrieved from SEC: https: //www.sec.gov/news/press fat-finge rs-cause -panic s/#180ad fe160 e9.-relea se/2020-5. Kashyap, K., and M.G. Weber. 2018. How Emerging Technologies will SEC and CFTC. 2010. Findings Regarding the Market Events of May Change Financial Services. In The FinTech Book, ed. S. Chisti and 6, 2010: Report of the Staffs of the CFTC and SEC to the Joint J. Barberis, 228. Cornwall: Wiley. Advisory Committee on Emerging Regulatory Issues. Washington, D.C.: SEC and CFTC. Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 335 SEC. 2010. Concept Release: Concept Release on Equity Market Struc-://www.frbat lanta .org/cenfi s/publi catio ns/notes fromt hevau lt/11- ture; Proposed Rule. Retrieved from SEC: https ://www.sec.gov/prudential -r egulation -bigda t a-and-machine-lear ning-2016-11-21. rules /conce pt/2010/34-61358 fr.pdf. Zeranski, S., and I.E. Sancak. 2020. Digitalisation of Financial Super- The World Bank. 2012. Rethinking the Role of the State in Finance: vision with Supervisory Technology (SupTech). Journal of Inter- Global Financial Development Report 2013. Washington, D.C.: national Banking Law and Regulation, pp. 309–330. The World Bank. The World Bank. 2020. Bank Regulation and Supervision a Decade Publisher’s Note Springer Nature remains neutral with regard to after the Global Financial Crisis: Global Financial Development jurisdictional claims in published maps and institutional affiliations. Report 2019–2020. Washington, D.C.: The World Bank. Wall, L.D. 2016. Prudential Regulation, Big Data, and Machine Learning. Retrieved from Federal Reserve Bank of Atlanta: https http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal of Disclosure and Governance Springer Journals http://www.deepdyve.com/lp/springer-journals/prudential-supervisory-disclosure-psd-with-supervisory-technology-JVHjFTr1Ph

Loading next page...

References (56)

A. Menkveld, B. Yueshen (2017)
The Flash Crash: A Cautionary Tale about Highly Fragmented Markets
Capital Markets: Market Microstructure eJournal
(2010)
SEC: No evidence that one event caused market plunge. Retrieved from Marketwatch
(2010)
SEC Proposes Consolidated Audit Trail System to Better Track Market Trades
(2013)
MIDAS-Market Information Data Analytics System
(2020)
Supervisory Disclosure
(2010)
Examining the Causes and Lessons of the May 6th Market Plunge
(2010)
Dow Takes a Harrowing 1 , 010 . 14 - Point Trip ; Biggest Point Fall , Before a Snapback ; Glitch Makes Thing Worse Definition of prudential in English
Findings Regarding the Market Events of
(2019)
Bank Regulation and Supervision a Decade after the Global Financial Crisis: Global Financial Development Report
(2010)
Stocks plunge as trading glitch suspected. Retrieved from Reuters
(2012)
Investor Bulletin: Measures to Address Market Volatility
A. Kirilenko, A. Kyle, M. Samadi, Tugkan Tuzun (2017)
The Flash Crash: High-Frequency Trading in an Electronic Market: The Flash Crash
Journal of Finance, 72
D. Easley, Marcos Prado, Maureen O'Hara (2011)
The Microstructure of the “Flash Crash”: Flow Toxicity, Liquidity Crashes, and the Probability of Informed Trading
The Journal of Portfolio Management, 37
D. Arner, J. Barberis, Ross Buckley (2016)
The Emergence of Regtech 2.0: From Know Your Customer to Know Your Data
Monetary Economics: International Financial Flows
A. bouveret, P. Breuer, Yiñgyuañ Cheñ, David Jones, Tsuyoshi Sasaki (2015)
Fragilities in the U.S. Treasury Market: Lessons from the 'Flash Rally' of October 15, 2014
Banking & Insurance eJournal
Yacine Ait-Sahalia, Mehmet Saglam (2013)
High Frequency Traders: Taking Advantage of Speed
ERN: Other Microeconomics: General Equilibrium & Disequilibrium Models of Financial Markets (Topic)
(2018)
Report on the proposal for a directive of the European Parliament and of the Council on the prudential supervision of investment firms and amending Directives 2013/36/EU and 2014/65/EU
P. Gomber, Jascha-Alexander Koch, Michael Siering (2017)
Digital Finance and FinTech: current research and future research directions
Journal of Business Economics, 87
(2010)
Testimony Concerning the Severe Market Disruption on May 6, 2010
T. Tressel, E. Detragiache, Asli Demirgüç-Kunt (2006)
Banking on the Principles: Compliance with Basel Core Principles and Bank Soundness
Banking & Financial Institutions
(2020)
EBA Response: EC consultation on the Digital Finance Strategy/Action Plan
(2017)
RegTech and SupTech: What They Mean for Financial Supervision
C. Borch (2016)
High-frequency trading, algorithmic finance and the Flash Crash: reflections on eventalization
Economy and Society, 45
(2020)
Digital Finance Series: Finance 2025
(2010)
Concept Release: Concept Release on Equity Market Structure; Proposed Rule
A. Kirilenko, A. Kyle, M. Samadi, Tugkan Tuzun (2018)
Automation, Intermediation and the Flash Crash
ERN: Futures (Topic)
(2010)
Fat Fingers Cause Panics
M. Kashyap, G. Weber (2016)
How Emerging Technologies Will Change Financial Services
D. Anginer, A. Demirguc-Kunt, D. Mare (2018)
Bank capital, institutional environment and systemic stability
Journal of Financial Stability, 37
Joel Peress (2011)
The Media and the Diffusion of Information in Financial Markets: Evidence from Newspaper Strikes
S&P Global Market Intelligence Research Paper Series
(2018)
The EBA’s FinTech Roadmap: Concluions from the Consultation on the EBA’s Approach to Financial Technology (FinTech)
(2019)
SEC Approves Plan to Create Consolidated Audit Trail
Dirk Broeders, Jermy Prenio (2018)
Innovative technology in financial supervision (suptech): The experience of early users
C. Goodhart (2002)
Prudential Supervision: What Works and What Doesn't
The Economic Journal, 112
(2020)
SEC Proposes Improvements to Governance of Market Data Plans
(2010)
FLASH CRASH May
(2018)
Supervising Tomorrow
(2016)
Prudential Regulation, Big Data, and Machine Learning
Stefan Zeranski, Ibrahim Sancak (2020)
Digitalisation of Financial Supervision with Supervisory Technology (SupTech)
Information Systems & Economics eJournal
(2016)
10 Key Marketing Trends for 2017
(2014)
Four - year flash crash anniversary haunts markets
(2018)
Big Data Meets Artificial Intelligence: Challenges and Implications for the Supervision and Regulation of Financial Services
Anna Klimina (2017)
Rethinking the role of the state
2020. Fintech in Europe and Central Asia: Maximizing Benefits and Managing Risks
Gunhild Berg, Mario Guadamillas, A. Sarkar, H. Natarajan (2020)
Fintech in Europe and Central Asia
(2020)
Transparency Requirements for Listed Companies
(2020)
catnmsplan
Alfons Bora, H. Hausendorf (2004)
Final Report to the European Commission
SEC Approves Vast Surveillance System for Stock Market
(1998)
Enhancing Bank Transparency: Public Disclosure and Supervisory Information that Promote Safety and Soundness in Banking Systems
(2019)
Global Financial Development Report 2019/2020: Bank Regulation and Supervision a Decade after the Global Financial Crisis
(2020)
When Volatility Surges, SEC’s Trade-Monitoring System Has Struggled
(2010)
Vital lessons of the ‘flash crash
(2017)
European Commission. 2020. Prudential Requirements
(2017)
A Quick Guide to Macroprudential Policies
Frederic Mishkin (2001)
Prudential Supervision: What Works and What Doesn't

Publisher: Springer Journals
Copyright: Copyright © The Author(s) 2021
ISSN: 1741-3591
eISSN: 1746-6539
DOI: 10.1057/s41310-021-00111-7
Publisher site: See Article on Publisher Site

Abstract

The U.S. financial markets faced an unprecedented rapid decline and recovery on May 6, 2010, known as the May 6 flash crash. Roughly one trillion $ market value in less than thirty minutes vanished with the biggest one-day point decline in the history of the DJIA at the time. Since the market events took place in electronic markets, and algorithmic trading and high- frequency trading, parts of FinTech, played significant roles, we handle the May 6 flash crash from the FinTech, SupTech, and financial supervision perspectives. With the flashback method, we analyzed the reactions of market participants, media, and two financial supervisors, the SEC, and the CFTC, to the market crash. We find that the technological imbalance between financial markets or institutions and their supervisors drove the markets in uncertainty, hence in a fear and panic environment. Since the imbalance has not diminished yet, the same risks still exist. As a remedy, we introduce a new concept and model with a well-functioning SupTech system to cope with the May 6 type FinTech crises. Keywords Supervisory technology · SupTech · FinTech · RegTech · Financial supervision · Financial system · Prudential supervisory disclosure · Financial authority · Digital finance · May 6 flash crash · FinTech crises · Financial crises · Financial stability · Informational efficiency · Systemically important data · Systemically important markets · Know-your- data · Know-your-technology · Know-your-markets · Inform-your-markets JEL Classification D47 · D53 · G18 · G01 · G28 · H11 · K22 · K23 · L15 · O31 · O32 Abbreviations CTS Consolidated Tape System AI Artificial Intelligence DJIA Dow Jones Industrial Average AT Algorithmic Trading EBA European Banking Authority BaFin German Federal Financial Supervi-EC European Commission sory Authority ECB European Central Bank BCBS Basel Committee on Banking EDGAR U.S. Electronic Data Gathering, Supervision Analysis, and Retrieval System BIS Bank for International Settlements EFIF European Forum for Innovation CAT Consolidated Audit Trail Facilitators CMB Capital Markets Board of Turkey EIOPA European Insurance and Occupa- CFTC U.S. Commodity Futures Trading tional Pensions Authority Commission E-Mini S&P 500 Futures Contracts CME Chicago Mercantile Exchange ESMA European Securities and Markets CQS Consolidated Quotation System Authority ETF Exchange Traded Fund EU European Union * Ibrahim E. Sancak EWS Early Warning System i.sancak@ostfalia.de FED U.S. Federal Reserve System Stefan Zeranski FDIC Federal Deposit Insurance st.zeranski@ostfalia.de Corporation Ostfalia University of Applied Sciences, Wolfenbüttel, Germany Vol.:(0123456789) 316 S. Zeranski, I. E. Sancak FINRA U.S. Financial Industry Regulatory Since the October 1987 crash originated at the U.S. mar- Authority kets, the financial market structure has evolved as techno- FinTech Financial Technology logical advancements that have enabled participants to trade FSB Financial Stability Board using algorithms with little or no human intervention (Kir- FSP Financial Services Provider ilenko et al. 2018). Today, digital finance and the FinTech HFT High-frequency Trading world have a lot of tools and technologies, including high- HFTs High-frequency Traders frequency trading (HFT) and algorithmic trading (AT). HFT IMF International Monetary Fund and AT were intensively debated due to the May 6, 2010, LRPs Liquidity Replenishment Points flash crash and received close attention by the public and MIDAS Market Information Data Analytics regulators (Gomber et al. 2017). This debate also triggered System intensive academic research on the impact of high-frequency NBBO National Best Bid and Offer trading and algorithmic trading on market quality, especially NLP Natural Language Processing market stability and integrity (Gomber et al. 2017). How- NMS National Market System ever, the roles of supervisors in the digital financial world NYSE New York Stock Exchange have not been discussed enough to develop new instruments OCC Office of the Comptroller of the to answer new risks arising from new technological tools, Currency market speed, and big data. The May 6 flash crash is a con- PSD Prudential Supervisory Disclosure venient case to analyze the roles and possible responding RegTech Regulatory Technology technologies and policies against FinTech-related risks. SEC U.S. Securities and Exchange The market events of May 6, 2010, shook the confidence Commission of market participants and raised questions about the mar- SPY S&P 500 SPDR Exchange Traded ket structure of electronic markets (Kirilenko et al. 2017). Fund (SPDR: Standard and Poor’s Considering the regulatory and supervisory responsibilities, Depository Receipt) it also raised questions about possible responses with SupT- SRO Self-Regulatory Organization ech. Most academic papers and discussions about the May SSM Single Supervisory Mechanism 6 flash crash have focused on the market microstructure. SupTech Supervisory Technology However, supervisory agencies’ roles are not less impor- TECHs in Finance FinTech, RegTech, SupTech, and tant than that. We believe that the May 6 case has more other “Tech” areas in Finance and at issues and implications than the concerns about the market Financial Sectors. structure. For example, the May 6 case addresses a market U.S. United States of America disorder that nurtures fear and panic and feeds market-made VIX Volatility Index detrimental stories about the crash, which might fuel further VPIN Volume-Synchronized Probability of crashes and vicious circles, ultimately crises. Informed Trading One of the concerns about the May 6 case was uncertainty during the market crash. With a flashback perspective, we collected information about responses of market participants Introduction and media to the crash. No one, including the supervisors, did know what happened during the day, even after the day Technological imbalance or asymmetric technology between until the supervisory agencies revealed a joint report with financial markets or institutions and their supervisors is convincing findings, approximately five months later, on more dangerous than cyber-attack risks since cyberattacks September 30, 2010. are well-known risk types; hence, there is considerable vigi- The SEC’s chairperson announced with a written state- lance to develop shields against them. However, the lack ment on May 20, 2010, that: "On the Monday following the of a well-functioning supervisory technology (SupTech) events of May 6, I met here in Washington with the lead- leaves many doors wide-open for detrimental technologi- ers of six markets—New York Stock Exchange, NASDAQ cal transactions and their ensuing effects on an economy’s Stock Market, BATS Exchange, Direct Edge ECN, Inter- financial stability. In other words, asymmetric technology national Securities Exchange, and Chicago Board Options between financial markets and the relevant supervisors is Exchange—and FINRA, to discuss the causes of market one of the most significant risks today. Therefore, having a events of May 6, the potential contributing factors, and pos- digital financial supervisory system with a well-functioning sible market reforms. The meeting was productive and col- SupTech is one of the best risk management strategies in laborative, and there was a strong consensus that the type this regard. of aberrational volatility experienced on May 6 is wholly Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 317 Fig. 1 Decline of E-Mini and DJIA Based on 11:00 AM Lev- els. Source: Schapiro (2010); Bloomberg unacceptable in our markets." The May 6 case has forced debt crisis, led to growing uncertainty in the financial mar - the U.S. financial supervisors to reform their infrastructure. kets (CFTC and SEC 2010). This negative sentiment-driven Analyzing the May 6 case and examining the reform sell pressure and flight to quality transactions accelerated the efforts, we have realized that financial supervisors need overall decline in the financial markets suddenly beginning entirely new instruments to monitor and supervise today’s shortly after 2:30 PM (CFTC and SEC 2010). markets. This paper mainly focuses on prudential supervi- The following graph indicates the path and the timing of sory disclosure and the reasoning behind it with a real and the E-Mini and the DJIA movements on May 6, based on stunning case, the May 6 flash crash. 11:00 AM levels (Fig. 1). Prudential supervisory disclosure is the set of disclosure As shown by the graph, the market crash deepened in rules for supervisors to inform the market participants timely 30 min, but market indicators came back in one hour, with about market-wide harmful conditions and activities to pre- around 2% decline based on 11:00 AM levels. serve market integrity and protect markets from detrimen- The following table is the flashback of the events of May tal rumors, orders and transactions by using their SupTech 6 (Table 1). capacity. This paper’s concept is much more related to the On May 6, in the four-and-one-half minutes (from 2:41 biggest economies, such as the U.S. and the EU, since their p.m. through 2:45:27 PM), prices of the E-Mini had fallen state-based or member-based economic areas are more con- by more than 5%, and prices of the SPY suffered a decline ducive to the PSD model. of over 6% (SEC and CFTC 2010). Internal stabilizers or This paper consists of two sections. In the first section circuit breakers carried out an important function to pre- of this paper, we analyze the May 6 market crash from the vent further liquidity and price collapses. For example, at supervisory technology perspective and the media and mar- 2:45:28 PM, trading on the E-Mini was paused for five sec- ket participants’ reactions. The second section explains the onds when the Chicago Mercantile Exchange (CME) Stop general framework of the new concept, prudential supervi- Logic Functionality was triggered; hence, the sell-side pres- sory disclosure, and its implications for banking and capital sure in the E-Mini was partly alleviated, and the buy-side markets sectors. The conclusion gives a summary of our interest increased, and when trading resumed at 2:45:33 PM, findings, views, and highlights. prices stabilized and shortly after that, the E-Mini began to recover, followed by the SPY (SEC and CFTC 2010). We Analysis of the May 6 market crash Stop Logic Functionality pauses trading when the trading engine and aftermaths recognizes that it has a series of resting stop orders that could lead to a cascade and move the market up or down beyond a specified amount (SEC and CFTC 2010). Chronological order of the May 6 market crash In contrast to the fact that the Chicago Mercantile Exchange (CME) Stop Logic Functionality was triggered and functioned, the SEC and On May 6, 2010, the negative sentiment, unsettling political the CFTC (2010) reported that the staffs of the CFTC and SEC were and economic news from overseas concerning the European working together with the markets to consider recalibrating the exist- 318 S. Zeranski, I. E. Sancak Table 1 The May 6 Market Crash Timeline Phases Time Event Effects and Results Phase I 1:00 PM Due to the news from overseas concerning the The number of volatility pauses, also known as European debt crisis, broadly negative market Liquidity Replenishment Points (LRPs), trig- sentiment was affecting an increase in the price gered on the New York Stock Exchange (NYSE) volatility of some individual securities in individual equities listed and traded on that exchange began to substantially increase above average levels 2:30 PM The S&P 500 volatility index (VIX) was up 22.5 Yields of ten-year Treasuries fell as investors percent from the opening level engaged in a “flight to quality,” and selling pressure had pushed the Dow Jones Industrial Average (DJIA) down about 2.5% During the first phase, from the open through about 2:32 PM, prices were broadly declining across markets, with stock market index products sustaining losses of about 3% Phase 2:32 PM A large fundamental trader (a mutual fund com- Sell pressure was initially absorbed by: II plex) initiated a sell program to sell a total of high-frequency traders (HFTs) and other interme- 75,000 E-Mini contracts (valued at approxi- diaries in the futures market, mately $4.1 billion) as a hedge to an existing fundamental buyers in the futures market, and equity position via an automated execution cross-market arbitrageurs who transferred this sell algorithm and executed the sell program by pressure to the equities markets by opportun- only targeting trading volume, and neither price istically buying E-Mini contracts and simul- nor time, extremely rapidly in just 20 min taneously selling products like SPY or selling individual equities in the S&P 500 Index As a result, HFTs accumulated a net long position of about 3,300 contracts From about 2:32 PM through about 2:41 PM, the broad markets began to lose more ground Phase 2:41 PM- 2:44 PM HFTs aggressively sold about 2,000 E-Mini con- Two liquidity crises – one at the broad index III tracts to reduce their temporary long positions. level in the E-Mini, the other with respect to At the same time, HFTs traded nearly 140,000 individual stocks E-Mini contracts or over 33% of the total trad- ing volume 2:45:28 PM Trading on the E-Mini was paused for five sec- Sell-side pressure in the E-Mini was partly allevi- onds when the Chicago Mercantile Exchange ated and buy-side interest increased Stop Logic Functionality was triggered to prevent a cascade of further price declines Volume spiked upwards and the broad markets plummeted a further 5–6% to reach intra-day lows of 9–10% Phase 2:45:33 PM-3:00 PM Trading resumed Prices stabilized and shortly thereafter, the E-Mini IV began to recover, followed by the SPY Broad market indices recovered while at the same time many individual securities and ETFs experienced extreme price fluctuations and traded in a disorderly fashion Phase 3:00 PM -Closings Most securities had reverted to trading at prices After the market closed, the exchanges and FINRA V reflecting true consensus values met and jointly agreed to cancel (or break) all such trades under their respective “clearly erro- neous” trade rules Prices of most individual securities significantly recovered, and trading resumed in a more orderly fashion Source CFTC and SEC (May 2010) and SEC and CFTC (2010); outlined by the Authors *Within the NYSE’s hybrid floor/electronic trading model, on May 6, the NYSE implemented price-bands known as “liquidity replenishment points.”, or LRPs. LRPs are intended to act as a “speed bump” and to dampen volatility in a given stock by temporarily converting from an auto- mated market to a manual auction market when a price movement of sufficient size is reached SEC and CFTC (2010) In the course of the day, VIX, a measure of the expected volatility of the S&P 500 Index, increased by 31.7 percent, which was the fourth largest single day increase in VIX SEC and CFTC (2010) It was realized that especially in times of significant volatility, high trading volume is not necessarily a reliable indicator of market liquidity SEC and CFTC (2010) contemplate that the circuit breakers played an important though the circuit breakers at different markets function with role in alleviating the crash stress and adverse effects. Even different parameters and protect the markets from broader and higher price and liquidity collapses, they are mainly not for public disclosure and an information source against Footnote 2 (continued) market rumors and uncertainty. Circuit breakers might be a ing market-wide circuit breakers—none of which were triggered on window for fresh air and an opportunity to gain some time May 6—that apply across all equity trading venues and the futures for making a prudential supervisory disclosure, but they markets. Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 319 cannot carry out any supervisory disclosure role. Even, cir- Our idea is that only exchanges or SROs like FINRA and cuit breakers might signal unintended messages to markets financial supervisors, such as the SEC and the CFTC, could and market participants. Moreover, there is no guarantee help protect the markets by making a statement from further that circuit breakers work properly or work at all in every deterioration. Uncertainty and the fear about the causes of case. During the U.S. Treasury “Flash Rally”, on October the May 6 crash harmed the market integrity and market 15, 2014, market safeguards could not prevent large price quality. The harm could be less with the supervisors’ imme- movements (Bouveret et al. 2015). diate disclosure statement in the sense of prudential super- Cross-market propagation issues were also in the center visory disclosure. However, at the time, neither the CFTC of the May 6 market crash. Since many products and mar- nor the SEC had the capacity to inform the markets with the kets are connected to each other, a market crash may trigger facts about the crash. another one, multiple exchanges and trading platforms might One of the observations is that many (though not all) be affected without knowing the real causes of the crashes firms significantly curtailed or completely halted their trad- they face at their markets. For example, a stock market crash ing activities at some point during the afternoon of May at an exchange might trigger other crashes at derivatives 6 (SEC and CFTC 2010). Data integrity issues were their exchanges due to the underlying stocks or stock indices. In number one concern. This also addresses the prudential this case, the derivatives exchange members or managers supervisory disclosure. With the real-time data collection cannot know the real causes of the market crash at deriva- capacity from multiple data centers and having data integ- tives markets. However, a supervisory agency that has a real- rity control capacity, during the cases like the May 6 market time data collection capacity from all markets, either stock crash, the supervisors can inform the markets and eliminate markets or derivatives markets, can capture the causes of potential concerns or clarify the situation for better market a market crash or intervene with before a crash comes out quality. by applying advanced data analytics. Thus, the May 6 type Data-integrity pauses were the reality of markets at the FinTech crises only can be managed by a real-time market- time. There were some other concerns about this unusual wide data collection capacity, which is one of the essential market behaviors. For example, at the time, it could be features of a well-functioning SupTech system. Moreover, hypothesized that these delays were due to a manipulative we should point out that the supervisory model of the U.S. practice called “quote stuffing” in which high volumes of financial markets is still not functional, since considering quotes were purposely sent to exchanges in order to create the case above, the SEC and the CFTC have different market data delays that would afford the firm sending these quotes responsibilities. Without perfect coordination and collabora- a trading advantage (SEC and CFTC 2010). tion, which is not a case in many times among national finan- Even though neither FinTech nor SupTech was a part of cial supervisors, cross-market propagation issues cannot be financial terminology at the time, the SEC and the CFTC’s managed, even with the real-time data collection capacity. report (September 2010) addresses FinTech, RegTech, and The May 6 market crash indicated that during the crash, SupTech concepts (TECHs in Finance) many times without not only media and individual investors but also institu- using these terms. The following paragraph is one of the tional investors did not know about the real causes of the key statements indicating agencies’ concerns in this regard: crash since they were not able to see the whole picture of “The events of May 6 clearly demonstrate the importance the markets. Considering asymmetric information envi- of data in today’s world of fully-automated trading strate- ronment during the market crash, based on their respec- gies and systems. The SEC staff will therefore be working tive individual risk assessments, some market makers and closely with the market centers to help ensure the integrity other liquidity providers widened their quote spreads, others and reliability of their data processes, especially those that reduced oe ff red liquidity, and a signic fi ant number withdrew involve the publication of trades and quotes to the consoli- completely from the markets (SEC and CFTC 2010). Not dated tape. In addition, the SEC staff will be working with only did some withdraw, but arguably they became liquidity the market centers in exploring their members’ trading prac- consumers by dumping their inventories, thus exacerbating tices to identify any unintentional or potentially abusive or the crash (Easley et al. 2011). The flash crash might have manipulative conduct that may cause such system delays that been avoided, or at least tempered, had liquidity providers remained in the marketplace (Easley et al. 2011). The paper of Easley et al. (2011) suggests the Volume-Synchro- nized Probability of Informed Trading (VPIN) as a solution. The VPIN might capture the increasing toxicity of the order flow in the Even though the market safeguards, circuit breakers, were not trig- hours and days prior. The VPIN contract might be used with pru- gered on October 15, 2014, they were triggered previously (Bouveret dential supervisory disclosure to monitor and manage similar risks et al. 2015). dynamically. 320 S. Zeranski, I. E. Sancak inhibit the ability of market participants to engage in a fair cannot completely rule out these possibilities.” This state- and orderly process of price discovery.” ment and the preliminary report’s general findings suggest Being supervisors of the foremost capital markets in the that both agencies did not have enough evidence to rule out world, the SEC and the CFTC have been aware of the impor- rumors and detrimental news feeds. The SEC’s chairperson tance of technology. However, digital transformation with also touched on the “fat finger” rumors as follows (Schap- a cutting edge SupTech system and a country-wide mar- iro, Testimony Concerning the Severe Market Disruption on ket reform requires political leadership and strong financial May 6, 2010, 11 May 2010): “There have been reports in the support, which have been not entirely in the hands of the press about a “fat-finger” error where, it is hypothesized, agencies. The same bottleneck holds for almost all financial an order of billions of shares was entered, rather than an regulators and supervisors. intended order of millions of shares. While we cannot yet definitely rule that possibility out, neither our review nor Reactions of the market participants and the media reviews by the relevant exchanges and market participants have uncovered such an error.” The reactions of market participants during the day on May The May 6 market crash also attracted the attention of the 6 have been analyzed by both financial agencies and academ - media. Economic and financial media are important sources ics. The SEC and CFTC reports tried to capture the behav- of information for financial consumers. Peress (2014) dem- iors of different market actors. In response to the increased onstrates that the media influence the stock market by risk perceptions, some market makers and other liquid- increasing the speed with which information diffuses across ity providers widened their quote spreads, others reduced investors and is impounded into stock prices. offered liquidity, and a significant number withdrew com- Borch (2017) evaluates the experimental impact (the pletely from the markets (SEC and CFTC 2010). The inves- Flash Crash’s effect on market participants), the real eco- tigations of the staffs of the SEC and the CFTC revealed that nomic impact (the actual economic effects) and the potential the largest and most erratic price moves observed on May 6 systemic impact (the systemic risk involved in algorithmic were caused by withdrawals of liquidity and the subsequent trading, as illustrated in the crash), and subsequently argues execution of trades at stub quotes. that each impact is contestable. Academic papers also analyzed the market structure and Borch (2017), inter alia, discusses the following three reached some results for the FinTech world. For example, aspects of the May 6 flash crash for today’s financial mar - the papers of Kirilenko, Kyle, Samadi, & Tuzun, The Flash kets: (a) as an event that significantly changed how market Crash: High-Frequency Trading in an Electronic Market participants perceive markets; (b) as an event that generated (2017) and Automation, Intermediation and the Flash Crash a massive loss of value, and hence had or could have had (2018) assert that HFTs behave differently than traditional considerable economic effects; and/or (c) as an event that is market makers; their behavior is empirically more consistent symptomatic of a novel set of systemic risks associated with with quote sniping than traditional market-making. algorithmic finance. The final report of the May 6 case was published on Sep- We took snapshots of some available news feeds about tember 30, 2010, and it has more robust evidence about the the a fl sh crash to conceptualize the ee ff cts of the uninformed causes of the events. However, with the preliminary report, media risk (Table 2). the CFTC and SEC’s staffs were considering some working The executive vice president and head of operations at the hypotheses. There were no clear and definite findings even NYSE Euronext’s New York Stock Exchange commented on though the agencies delivered 150-page preliminary reports the May 6 crash that “This highlights the risks of electronic on May 18, 2010, twelve days later. trading. When you have low volatility, electronic trading For the informational efficiency and market integrity works very well. But there are risks. It highlights the need for concerns, market participants should be informed by the human-based intervention.” (Lauricella and McKay 2010). responsible authorities in critical times and should not be Four years later, a news portal commented on the case as left in the hands of rumor feeds. We also believe that mar- follows (CNBC 2014): ket participants should be well informed to reduce adverse The Dow Jones Industrial Average slumped nearly selection risks and other outcomes of the inefficient infor - 1000 points in a matter of minutes in the flash crash of mational environment. 2010, sending traders into a panic and inciting scrutiny Some financial news sources were mentioning about a of the U.S. equities markets that’s still being felt four “fat finger” issue as the triggering source of the crash. The years later. preliminary report’s response to these rumors and news The May 6, 2010, crash was initially blamed on a “fat- was as follows (CFTC and SEC 2010): “We have found no finger” error made at Citigroup-a theory that was later evidence that these events were triggered by “fat finger” shot down and ultimately attributed to investment firm errors, computer hacking, or terrorist activity, although we Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 321 Table 2 Economic and Financial Media Comments on the Market Events of May 6, 2010 News source Date Comment Reuters 6 May 2010 “The Dow suffered its biggest ever intraday point drop —998.5 points. The market’s fall may have been exacerbated by erroneous trades that showed some shares briefly fell to nearly zero The situation remained unclear long after the closing bell as the Nasdaq Stock Market and others said they would cancel multiple erroneous trades. Other exchanges scrambled to examine orders.” (Krudy 2010) Forbes.com 6 May 2010 “Why the market plunged so much and so fast in the middle of the afternoon isn’t entirely clear. Some blame an erroneous quote on Procter & Gamble (PG), saying it caused panic selling across the board. Others say the selloff was caused by a trading error on the Nasdaq. This so-called fat-finger error occurred when a trader accidentally entered an order to sell a billion shares rather than a million shares. Still others blame the rioting in Greece for the selloff. That rioting was widely broadcast on trading floors." (Janjigian, 2010) CNBC.com (TV) 6 May 2010 A commentator on TV: “…machines broke down…”, “…the system obviously broke down…” “…it broke down, machines broke down…” A speaker on TV: “…there should be an investigation…” A speaker asks a question about the P&G prices: “…with P&G is there any rational way that you could describe P&G being three percent down and then 25 within what was in 90 s to three minutes?…” A P&G analyst answers: “…no as machines that to be broken I mean there is no fundamental reason for Procter to be down more than two percent today…” (CNBC 2010) Wall Street Journal 7 May 2010 “A bad day in the financial markets was made worse by an apparent trading glitch, leaving traders and inves- tors nervous and scratching their heads over how a mistake could send the Dow Jones Industrial Average into a 1,000-point tailspin.” “Traders theorized that an initial trading error triggered a piling-on effect from computerized trading programs designed to sell when the market moves lower. At the same time, pre-set orders from individual traders and investors to sell on declines during market downturns were likely triggered.” “The move highlighted how fragile U.S. markets have become and how the various fragmented markets have deficiencies in the way they buffer volatility.” (Lauricella and McKay 2010) Marketwatch.com 11 May 2010 “Two top financial regulators said Tuesday they aren’t sure yet what caused the stock market’s dizzying May 6 plunge and partial recovery, but they don’t believe any one event created it At issue is the Dow Jones Industrial Average drop of nearly 1,000 points last Thursday—a fall of roughly $1 trillion in market value—much of it in a matter of minutes, before recovering to a 348-point loss for the session Both Mary Schapiro, Securities and Exchange Commission chairwoman, and Gary Gensler, Commodity Futures Trading Commission chairman, refuted speculation that a trader might have made a so-called "fat finger" error that contributed to the stock market plunge.” (Orol 2010) Financial Times 14 May 2010 “Just over a week later the cause of the “flash crash”, where in the space of those 20 min stocks plunged and rebounded, still remains a mystery. Talk, however, circulates that an algorithm, or “algo” computer program that dominates trading these days, may have exploited an already nervous market by sinking the shares and then buying them back at much cheaper prices.” (Mackenzie 2010) Source References and the Authors Waddell & Reed. But in addition to that trading error, units did have a data-driven explanation about the market a number of possible reasons for the crash has since turmoil. On May 6, traders were also stunned by the sudden come to light. One of those supposed causes was high- sharp moves (Lauricella and McKay 2010). frequency trading, according to a report from the Secu- Risk and uncertainty are entirely different concepts. rities and Exchange Commission that year. Uncertainty leaves risk management techniques ineffec - tive. For financial markets, uncertainty also fuels rumors As another comment made in the fourth anniversary of the and home-made stories about market events, as the markets flash crash has been raised an idea that the causes are still experienced during and after the May 6 market crash. The not fully agreed (Krantz 2014): best strategy in these cases is getting rid of uncertainty. In Even four years after the crash that wiped out $1 tril- this regard, financial supervisors should have an automated lion in wealth in the blink of an eye, investors and and real-time data collection capacity with advanced data academics still haven’t agreed on what caused one of analytics tools as well as prudential supervisory disclosure the most vicious and inexplicable short circuiting of capacity to keep the markets running without uncertainty. market to occur. On May 6, the market participants were in a panic situa- tion, and neither market participants nor market surveillance 322 S. Zeranski, I. E. Sancak fully automated data collection system (CFTC and SEC Supervisory technological capacity of the U.S. financial authorities 2010). The May 6 market crash also an indicative market event On the other hand, there was no evidence that both agen- cies deployed advanced data analytics technologies such as in terms of the technological capacity of the U.S. financial authorities in 2010. There were and still are two main regu- artificial intelligence, machine learning, natural language processing. The above picture was as of May 2010. How- latory and supervisory agencies for the U. S. capital markets: The Securities and Exchange Commission, the SEC, and ever, Broeders and Prenio (2018) mentions data analytics tools that the SEC either uses or projects to use as of 2018. the Commodity Futures Trading Commission, the CFTC. Two agencies established a Joint CFTC-SEC Advisory Com- In response to the situation realized after the May 6 flash crash, the SEC announced a project that may eliminate leg- mittee on Emerging Regulatory Issues. The Committee’s establishment was one of twenty recommendations included acy systems, let the agency collect real-time basis data, and deploy advanced data analytics. We will handle the policy in the agencies’ joint harmonization report issued in October 2009 (CFTC and SEC 2010). The joint committee published response of the U.S. financial supervisors in the following chapters. However, the May 6 case clearly indicated that the the “Preliminary Findings Regarding the Market Events of May 6, 2010,” on May 18, 2010, twelve days after the case. U.S. financial supervisors, namely the SEC and the CFTC did not have good enough SupTech capacity at the time, even The report’s content draws a general picture about the ini- tial findings of the May 6 market crash. The same report though the U.S. is one of the leading technology innovating countries in the world. also gives the framework of the U.S. financial markets as well as the supervisory technology at the time. The follow- Since the October 1987 crash, the market structure has evolved as technological advancements have enabled par- ing statement with the CFTC and the SEC (2010) report is a summary of the SupTech capacity of the U.S. financial ticipants to trade using algorithms with little or no human intervention (Kirilenko et al. 2018). The requirements of supervisors at the time: supervisory technology for the U.S. markets have been sig- It is important to emphasize that the review of the naled since the 1980s; however, the pace of the technol- events of May 6 is in its preliminary stages and is ogy adoption has been significantly different between the ongoing. The reconstruction of even a few hours of markets and their supervisors. We observe here another fact trading during an extremely active trading day in mar- that it is not about having available technology; it is about kets as broad and complex as ours— involving thou- organizing, designing, and having a well-functioning SupT- sands of products, millions of trades and hundreds of ech and supervisory system at large. On the other hand, as millions of data points—is an enormous undertaking. we pointed out in different sections, having a full-fledged Although trading now occurs in microseconds, the SupTech system is not entirely tied to financial supervisors. framework and processes for creating, formatting, and It requires additional funds and political support as well as collecting data across various types of market partici- leadership. pants, products and trading venues is neither stand- Since the U.S. capital markets have more than one super- ardized nor fully automated. Once collected, this data visor, authorities shared the workload of analyzing the data must be carefully validated and analyzed. Such further considering their responsibility areas. In this regard, for data and analysis may substantially alter the prelimi- example, the SEC has sourced and analyzed price, time, nary findings presented in this report. The staffs of and volume data on over 19 billion shares executed on May the Commissions therefore expect to supplement this 6, and quote data representing the best bid and best offer report with further additional findings and analyses. for over 7,800 securities, for each exchange, for each mil- In summary, the U.S. financial authorities, the SEC and lisecond during the trading day, and the CFTC has analyzed transaction and order book data on stock index futures, the CFTC, in 2010, did not have. including the E-Mini S&P 500 futures contract (CFTC and • SEC 2010). Data collection, consolidation, and data analyt- real-time data collection capacity, • cross-market surveillance capacity, ics from two different channels by two different supervisors • in the same jurisdiction for the same case are not effective consolidated transaction data collection capacity, • consolidated order tracking system, supervisory strategies in the FinTech world. • After the May 6 case, the SEC emphasized the impor- standardized data, tance of having a consolidated order tracking system or consolidated audit trail system. If adopted, this rule pro- posal should result in a continuous reporting mechanism for 5 –6 market participants to capture the data needed for effective One million microseconds are equal to one second (1 µs = 10 s). Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 323 Table 3 Overall quality check of the dimensions of the U.S. financial to change this picture in a free market economy. Second, supervisory system as of May 2010 the supervisory structure is fragmented, and this also poses a significant risk for the markets. The May 6 case forced to Feature Status bring the CFTC and the SEC together on a project basis. Organizational Structure Fragmented However, later, the flash events in the U.S. Treasury markets Supervisory Model Not for the of 15 October 2014 led to a bigger coordination require- FinTech ment: the U.S. Department of the Treasury, the FED, the World Federal Reserve Bank of New York, the SEC and the CFTC. Real-Time Data Collection NA Restructuring economic and financial agencies is a national Automated Data Collection Partly economic area, and it is in the hands of the governments and Digital Identification NA politicians. It is a strategic risk management area. Thus, it is Early Warning System NA a real challenge but still possible. Regulatory and Industry Sandboxes NA Data Analytics Partly Supervisors’ technological, administrative, Prudential Supervisory Disclosure NA and policy responses Source The Authors The exchanges report the daily positions and transactions of each The CFTC and SEC’s initial report, Preliminary Findings clearing member to the CFTC and the data are transmitted elec- Regarding the Market Events of May 6, 2010, was partially tronically during the morning after the “as of” date SEC and CFTC relieving work to eliminate some rumors and cascading (2010). The CFTC also collects trade data on a daily, transaction date + 1 (“T + 1”), basis from all U.S. futures exchanges through effects of the crash; however, it was not a clear answer at Trade Capture Reports SEC and CFTC (2010) the time. The report stated the lack of. For the CFTC, all transactional data is received overnight, loaded in the CFTC’s databases, and processed by specialized software applica- real-time, tions that detect patterns of potentially abusive trades and alerts SEC standardized, and CFTC (2010) automated data collection features. cross-market surveillance (CFTC and SEC 2010). Cross- These features are the main pillars of digital financial market surveillance was not in play on May 6. Therefore, it systems in the FinTech world (Zeranski and Sancak 2020). was impossible to capture the big picture of the U.S. capi- The content of the initial report and some statements, such tal markets. Together with other lessons, the May 6 crash as “It is important to emphasize that the review of the events was an important reminder of the inter-connectedness of of May 6 is in its preliminary stages and is ongoing.” and derivatives and securities markets, particularly with respect “Much work is needed to determine all of the causes of the to index products (SEC and CFTC 2010). market disruption on May 6.”, were not so helpful to man- As specified with the paper (Zeranski and Sancak 2020), age the fear of the markets. The report’s main message was one of the most critical features of a financial supervisory not signaling a strong perception of the technology-oriented system is the real-time data collection. For example, the supervision of the markets. Instead, it was signaling that exchanges report daily to the CFCT, and the agency con- there was a considerable gap between the technology that ducts daily surveillance with them. Daily surveillance seems markets used and the supervisors’ technology. In other to be a close look at the market; however, it is not enough in words, asymmetric technology was the case. What makes the FinTech environment. the case worse, the U.S. financial markets faced an undefined Considering the main features of a digital financial system situation, and the supervisors were not in the capacity to spot set forth by the paper Zeranski and Sancak (2020), we evalu- the causes of the crash until the final report came out on Sep- ate the overall quality of the dimensions of the U.S. financial tember 30, 2010. The reform efforts and official statements supervisory system with the following table (Table 3). also signaled a long way to close the gap. Beyond the table above, the U.S. financial markets have Knowing the root causes of a market crash enables two structural weaknesses from the supervisory perspective. supervisors to respond to the drivers of the crash timely and First, the U.S. markets are fragmented. However, it is hard adequately. The May 6 case indicates that supervisors did not have SupTech tools to respond to the drivers or decide whether any additional supervisory measures should be Although fragmented markets may have many implications, in the context of the May 6 crash, Albert J. Menkveld and Bart Zhou taken on May 6 or in the aftermath. In other words, at the Yueshen’s research, The Flash Crash: A Cautionary Tale about time, neither financial supervisors nor market participants Highly Fragmented Markets, suggests that liquidity supply in severely did know what happened and why it happened exactly. Prob- fragmented markets might become vulnerable when liquidity is ably the large trader who gave momentum to the market demanded. 324 S. Zeranski, I. E. Sancak crash with its algorithmic trading also did not know what between trading venues for exchange-traded funds, happened and why it happened. equity index futures, and equity index options—instru- The May 6 case raised another question at the time: Who ments used by investors to manage their exposures in was responsible for the supervision of the May 6 market the face of broad market movements. crash? The SEC or the CFTC? Since the causes of the mar- We agree the idea stated with the CFTC and the SEC report ket crash were not known during the day, the responsible (2010) that a uniform circuit breaker rule, which would supervisor could be one of them or both of them could be. briefly pause trading across the securities markets when the Or, the worst-case scenario; none of the agencies assumes price of a security has rapidly declined over a short time, responsibility to act immediately. This question also points should make a recurrence of a severe market disruption, like out another fact: The organizational model of the U.S. finan- the one that occurred on May 6, much less likely. However, cial markets per se a source of risk in the sense that there circuit breakers neither inform financial supervisors nor mar - was an ambiguity about the responsible supervisory author- ket participants about exactly what happened at the market. ity to respond to the market crash. The responsible super- Moreover, circuit breakers may not function in every case. visor might be the CFTC or the SEC. Since there was no Designing a well-functioning circuit breaker system is a solid information about the causes of the flash crash, it was significant step to relieve extraordinary market movements. also unclear that which of them would act against the market However, it does not give a picture of abusive market trans- crash. Thus, the supervisory model of the U.S. financial sys- actions. If we assume that a circuit breaker system works tem has been improper for the FinTech world. The country’s very well, but the supervisors do not have real-time data model has been under discussion after the global financial collection capacity, in this case, supervisors still may not crisis of 2008. Today, the fast-developing FinTech sector know what happened at the market. As stated within the sec- addresses the need of reform requirements again. ond report (2010), market participants might interpret a trig- It can be said that coordination is the answer. However, gered circuit breaker with their own story and might attribute coordination between two different independent supervisory greater importance to the circuit breakers or market pauses. organizations is both an intricate issue and time-consuming Thus, we infer that particularly for fragmented and relatively in practice. In the FinTech world, market crashes require bigger financial markets, precisely for the U.S. capital mar - prompt response and reaction. There is no time to develop kets, a centralized responsible authority should inform the new formal working groups from different organizations for markets about the nature of the extraordinary market events. taking prompt actions during market crashes like the May 6 We name this kind of announcements or disclosure policy crash. Thus, cumbersome, and difficult-to-update systems, as “prudential supervisory disclosure”. As we explain it in like the U.S. financial supervisory system, are relatively a separate section, an important feature of disclosure policy riskier than the easy-to-update financial systems. is that announcements are not discretionary. Thus, market Unless financial supervisors have multi-market and multi- participants know that there will be an announcement, and asset supervision capacity, having successful supervisory they will know what is happening exactly. infrastructure at one part of the markets is not enough for To sum up, whatever was the root cause of the flash protecting market integrity. The U.S. financial supervisors crash, market participants and the public should have been have been aware of this fact and their report (2010) put it informed about that. Unless an authority, which has a capac- in this way: ity to capture the picture of the market, announces the facts An important lesson from the events of May 6 is about unusual market events, market participants and media the need to better understand cross-market linkages will produce their stories. In this regard, financial super - visors should have a SupTech capacity such that they can capture all market activities, namely orders and transactions in a real-time basis, and in certain situations, public disclo- We assume that none of the traders has an intention to cause the sure should be mandatory but not discretionary for financial market crash at the time. Since no one, including financial supervi- supervisors so as not to let media and market participants sors, has the capacity to see all the data at all markets, we assume that a single trader also cannot know the causes of the May 6 crash during use their news production vision about probable causes of the day. a market event. In fact, it was turned out that both the equity and the derivatives On the other hand, both the SEC and the CFTC took les- markets experienced severe declines and disorders on May 6. Thus, sons from the May 6 case and started new projects against the May 6 case was about both the securities market and the deriv- similar crises. atives market. That means the case was both in the SEC and in the CFTC’s areas of responsibility. In order to increase the timeliness and efficiency of There is a misconception that developed countries have always account identification, the CFTC was considering possible advantage in terms of technological reforms. In fact, this is not the rules to enhance the CFTC’s surveillance capabilities by case for every developed country. Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 325 deploying automation of the statement of reporting traders The CAT will track orders throughout their life cycle in the large trader reporting system and obtaining account and identify the broker-dealers handling them, thus allow- ownership and control information in the exchange trade ing regulators to more efficiently track activity in Eligible registers (CFTC and SEC 2010). Securities throughout the U.S. markets (FINRA 2020). On May 20, the SEC’s chairperson stated that (Schapiro, Through the CAT, regulators in the U.S. expect to have more Examining the Causes and Lessons of the May 6th Market timely access to a comprehensive set of trading data, ena- Plunge, 20 May 2010); bling authorities to more efficiently and effectively conduct research, reconstruct market events, monitor market behav- During a 20-min period during the afternoon of May ior, and identify and investigate misconduct (SEC 2019). 6, the U.S. financial markets failed to live up to their The SEC estimates that the system will cost 2.4 billion USD essential price discovery function. That period of initially and then 1.7 billion USD a year to run (Bullock and gyrating prices directly harmed those investors who Stafford 2019). The CAT project works are still ongoing, and traded based on flawed price discovery signals, and it the project has not been in play yet. undermined the confidence of investors in the integrity On the way of having a well-function SupTech system, of the markets. We are committed to taking all neces- the SEC has outsourced the Market Information Data Ana- sary steps to identify causes and contributing factors lytics System (MIDAS). The history of MIDAS began with and are already working to reduce the likelihood of a the need to more efficiently collect and analyze order book recurrence of that day. data for equities and futures (SEC 2013). According to the Some of the SEC’s proposals were about the market struc- SEC, MIDAS has many applications at the SEC, and it can ture of the U.S. capital markets. Since the market structure help the agency monitor and understand mini-flash crashes, is not the main theme of this paper, we focus on other pro- reconstruct market events, and develop a better understand- posals, mainly ones about the supervisory capacity of the ing of long-term trends. However, there are some concerns agency. about the success of the system (Podkul 2020). One of the critical steps on the way of development of The agencies’ preliminary report stated that “Although a well-functioning SupTech system has been the Consoli- the coordinated circuit breakers between futures and equities dated Audit Trail (CAT) project. In this regard, the SEC’s were not triggered, the events of May 6 reinforce the impor- chairperson stated that (Schapiro, Examining the Causes and tance of having communication links between futures and Lessons of the May 6th Market Plunge 2010); equity markets so that there is meaningful and appropriate coordination of trading pauses and halts.” One of the challenges we face in recreating the events The SEC and the CFTC’s circuit breaker project might of May 6 is the reality that the technologies used for be a good fit for the prudential supervisory disclosure sys- market oversight and surveillance have not kept pace tem. As stated by the agencies within their September 2010 with the technology and trading patterns of the rapidly report, pausing a market might be an effective way of pro- evolving and expanding securities markets.” viding a window for market participants to reassess their “Today’s fast, electronic, and interconnected markets strategies, for algorithms to reset their parameters, and for demand a robust consolidated audit trail and execution an orderly market to be re-established. In this regard, the tracking system. CME’s Stop Logic Functionality helped prevent a possibly The SEC staff started in 2009 to work, in consultation bigger and more detrimental market crash by triggering a with SROs and others, on a rule proposal that would require halt in E-Mini trading. the SROs to jointly develop, implement and maintain a con- On May 31, 2012, the SEC approved a “Limit Up-Limit solidated order tracking system, or consolidated audit trail Down” mechanism to address market volatility by prevent- (Schapiro, Examining the Causes and Lessons of the May ing trades in listed equity securities when triggered by large, 6th Market Plunge 2010). The expectations with the CAT sudden price moves in an individual stock (SEC 2012). In project were mainly to increase the ability to access in real- July 2012, the SEC also announced that the securities and time the majority of the data needed to reconstruct the type futures exchanges have procedures for coordinated cross- of the May 6 market disruption, to enhance the ability to market trading halts if a severe market price decline reaches detect and monitor aberrant and illegal activity across mul- levels that may exhaust market liquidity (SEC 2012). These tiple markets. market-wide circuit breakers may halt trading temporarily or, under extreme circumstances, close the markets before the normal close of the trading session. During the crashes, some trades might be carried out 10 with erroneous prices and later might be broken or canceled. For a summary of the SEC’s proposals against the May 6 type However, market participants cannot exactly know which of market crashes, please see Schapiro (2010). 326 S. Zeranski, I. E. Sancak their trades will be canceled, and this uncertainty may cause a contract, as the uninformed party, runs the adverse selec- further trading problems and produce additional risks. For tion risk. Therefore, information disclosure is a fundamental example, market participants might not provide liquidity in ingredient of contracts and transactions. Public disclosure such a case. As seen this an important lesson taken from the has a similar function. In a public disclosure case, one of May 6 case, to provide market participants more certainty the parties informs the unknown people, makes announce- as to which trades will be broken and allow them to better ments to the public to maintain a level playing field. There manage their risks, the SEC staff worked with the exchanges are specially designated disclosure platforms to operate pub- and FINRA to clarify the process for breaking erroneous lic disclosure activities in financial markets. For example, trades using more objective standards (SEC and CFTC EDGAR is the web-based public disclosure platform for the 2010). By using real-time data infrastructure, it can be much U.S. capital markets. An Internet-based platform enables faster to spot erroneous data and manage the broken and all parties to reach the publicized information and help col- erroneous trades. lect data. Public disclosure is one of the main features of On the other hand, in January 2020, the SEC announced capital markets, particularly stock markets—public state- that the agency would modernize the national market sys- ments and filings flood markets with other market informa- tem. The SEC’s Chairman stated that “The Commission has tion. For example, the EDGAR’s system processes about received extensive public input on issues relating to equity 3,000 filings per day, serves up 3,000 terabytes of data to market structure and access to market data, as well as sug- the public annually and accommodates 40,000 new filers per gestions for how that structure should be updated to ensure year on average (SEC 2020a, b). Therefore, data visualiza- that our markets continue to best serve the interests of inves- tion, AI, NLP, and other technology tools are now required tors. Today’s proposed order is designed to address issues to benefit from available data and information. regarding the dissemination of market data that affect the Capital markets produce in every business day, on the one efficiency and fairness of our markets. In particular, we side, massive amounts of publicized information that are welcome public input on the specific proposed governance available at the public disclosure platforms, like EDGAR, provisions.” (SEC 2020a, b). This academic paper might be on the other side, generate order and transaction data in mar- a supportive work to help increase the efficiency and fair - kets. Data vendors sell these data to their clients. Therefore, ness of the U.S. markets. However, PSD is also conducive to not everybody has all the data outreach capacity. On the other markets, particularly complex markets with multiple other hand, many financial institutions submit data to the trading venues. In this regard, the European Commission supervisors under supervisory disclosure requirements. In should also think about PSD implementation, considering this world, only several organizations can legally see all data the member states’ markets. in the financial sector. Those are financial supervisors. Financial supervisors have macro-prudential responsibili- ties. “Prudential” is literally in the meaning of “involving or Prudential supervisory disclosure showing care and forethought, especially in business” (Lex- for the digital financial world ico 2020). “Prudence” is another word for caution involving forethought, and prudential policies relate to actions that Terminology: “Prudential” , “Supervisory Disclosure” , promote sound practices and limit risk-taking (European and “Prudential supervisory disclosure” Central Bank 2017). Prudential requirements aim at mak- ing the financial sector and economy sounder and more sta- Information disclosure, public disclosure, and supervisory ble. For example, the EU rules on prudential requirements disclosure are well-known concepts in finance and the finan- mainly concern the amount of capital and liquidity of banks cial sector. Information disclosure helps contractual parties (European Commission June 2020). For the banking sector, to know about all relevant information and facts of a trans- the goal of these rules is to strengthen the EU banking sec- action or a contract. For example, a bank informs its clients tor’s resilience so that it can better absorb economic shocks when the clients would like to buy products. If one party while ensuring that banks continue to finance economic does not inform the other party fully and causes asymmetric activity and growth (European Commission 2020). information between parties, multiple risks might arise. For example, if a party does not have the full information about EDGAR is the Electronic Data Gathering, Analysis, and Retrieval As stated by the SEC and CFTC’s report (2010), on September 10, system used at the SE (SEC 2020a, b). Containing millions of com- the SEC approved the new trade break procedures, which like the cir- pany and individual filings, EDGAR benefits investors, corporations, cuit breaker program, is in effect on a pilot basis through December and the U.S. economy overall by increasing the efficiency, transpar - 10, 2010. ency, and fairness of the securities markets (SEC 2020a, b). Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 327 Mishkin (2000) broadly defines prudential supervision world, the EU should consider the PSD model for the Union- as “government regulation and monitoring of the banking wide risk management mechanism against market-driven system to ensure its safety and soundness”. The forms of risks in addition to the investment firms-based prudential prudential supervision might be. supervision. The European Banking Authority (EBA) could play a restrictions on asset holdings and activities, leading role in networking to help initiate a well-function- separation of banking and other financial industries like ing SupTech system across the Single Market (EBA 2020). securities, insurance, or real estate, The EBA considers that the European Forum for Innovation restrictions on competition, Facilitators (EFIF) provides a good means for supervisors to capital requirements, share experiences on a cross-sectoral basis, aiding the iden- risk-based deposit insurance premiums, tification of innovation trends, regulatory and supervisory disclosure requirements, issues that require a cross-sectoral position and to moni- bank chartering, tor interconnectedness on a multi-disciplinary basis (EBA bank examination, 2020). supervisory versus regulatory approach (Mishkin 2000). There are supporting views of the banking sector of the EU that FinTech activities give rise to not only operational Mishkin’s approach to prudential supervision is mainly risks but also financial risks, especially of a systemic nature related to the banking sector. However, his approach does (EBA 2018). And, a potential ‘FinTech bubble’ was raised as not consider the speed and big data factors that markets face another issue of concern with the risk of reducing the effec- today. Two decades later, we have today completely different tiveness of monetary policy noted as another potential threat financial technology. Hence, prudential supervision can be (EBA 2018). Thus, the PSD model should be considered in used for many more areas and may have a broader mission response to FinTech risks. if we enter other avenues of the financial sector. Prudential supervision can be classified into micro- and Wall (2016) addresses advanced analytics and the impor- macro-prudential supervision. The prefix “macro” indicates tance of data by stating that the availability of more granular that the policies or actions relate to the whole or significant information, combined with new tools to analyze them, may parts of the financial system rather than individual financial provide supervisors with a variety of opportunities to evalu- institutions (European Central Bank 2017). Supervisory or ate the risk of financial systems better. The development of regulatory policies for individual financial institutions, by machine learning using deep learning techniques raises the contrast, are known as micro-prudential policies (European possibility that supervisors will be able to use granular data Central Bank 2017). While macro-prudential policies con- to better understand the risks in the financial system (Wall sider the soundness of the whole financial system, soundness 2016). and informational efficiency of markets should also be in this Sound prudential supervision policies should take into scope. However, today’s macro-prudential policies do not account the potential for investment firms and their clients focus on financial market-driven informational imbalances, to engage in excessive risk-taking and the different degrees particularly FinTech environment data production, usage, of risk assumed and posed by investment firms (European and the speed factors at the electronic markets. Parliament 2018). With the PSD model, we also raise the In other words, the FinTech world brings new responsi- idea that not only financial institutions but also markets bilities to financial regulators and supervisors, as financial should also be the realm of prudential supervision for the institutions have the responsibility of Know-Your-Customer, financial regulators and supervisors. According to a report financial supervisors should also have “Know-Your-Mar - of the European Parliament (2018), differences in the appli- kets”, “Know-Your-Technology”, “Know-Your-Data”, and cation of the existing framework in different member states “Inform-Your-Markets” responsibilities. The truly trans- of the EU threaten the level playing field for investment formative potential of regulatory technology addresses a firms within the Union, hampering investors’ access to new Know-Your-Customer mindset transformation into a Know- opportunities and better ways of managing their risks. This Your-Data approach (Arner et al. 2016). FinTech crises are also holds for technology and the market data or prudential a new source of systemic risks and the market data are sys- supervisory disclosure. temically important in the FinTech world. In this regard, all Since there might be differences in prudential supervisory market data should be considered as a new area for macro- capacities among the EU member states, creating a mecha- prudential supervision. nism of cooperation and exchange of information among Markets produce massive amounts of data every day, the financial authorities to ensure harmonized prudential and data analytics tools should be deployed at supervisory supervision of investment firms across the Union seems to agencies. Without data analytics tools and a well-designed be essential (European Parliament 2018). In the FinTech SupTech system, big data might be a black hole for financial 328 S. Zeranski, I. E. Sancak supervisors. By knowing the market data and conducting but supervisors do not have specific disclosure require- data analytics, financial supervisors can utilize big data ments regarding market events. For example, Turkey’s and even stay ahead of markets. The PSD model requires main supervisor for capital markets, the CMB, is not under a well-designed SupTech system. The PSD model has both a responsibility to inform the markets publicly under a writ- a systemic risk management capacity and implications for ten disclosure policy. There are some obscure and general transparency requirements. requirements, but there is not a specific requirement and a The growing FinTech world addresses new business mod- policy document or a guide to inform related parties at the els for private sector firms and new supervision models for Turkish capital market. Therefore, market participants do not financial supervisors. In this regard, based on the prudential exactly know when the CMB will inform them or whether supervisory disclosure (PSD), we introduce the PSD model. the CMB will make any announcement or not. In the FinTech world, supervisors should react promptly The scope, design, and functions of prudential to market events. Any delayed reaction might cause inevita- supervisory disclosure ble losses and market crashes. The May 6 market crash is the case indicating the timing concern in the FinTech world. On Supervisory disclosure or public disclosure are well-known May 6, when markets were already under stress, a sell algo- concepts in the financial industry or academic world. These rithm chosen by a large trader to only target trading vol- are regulatory requirements for market participants but not ume, and neither price nor time, executed the sell program for supervisors or financial authorities. extremely rapidly in just 20 min (SEC and CFTC 2010). Financial services providers are required to disclose some During the May 6 market crash in the U.S., there were many information either to supervisors or to the public. On the unknowns and fears of unknowns about probable roots of other hand, public disclosure is one of the main responsi- the crash, and supervisory agencies were under stress to act bilities of publicly held companies. There are strict rules against the drivers of the crash. They could not reveal useful for publicly held companies to reveal proper information information to the public timely. Six days after the crash, timely for related parties; shareholders, investors, and oth- the CFTC and the SEC published a report about the market ers. In all modern financial markets, public disclosure is one events. However, the report did not exactly answer the ques- of the main features and regulated areas. For example, the tions. Twenty days later, on May 26, 2010, the chairperson Transparency Directive (2004/1009/EC) requires issuers of of the SEC made a statement and pointed out that the SEC, securities traded on regulated markets within the EU to make at the time, could not track data across multiple markets, their activities transparent, by regularly publicizing certain products, and participants in a real-time basis. Later, to fill information (European Commission 2020). In this regard, the gap, the SEC introduced a Consolidated Audit Trail pro- the information to be publicized includes (European Com- ject or CAT in short. The following statement of the chair- mission 2020): person of the SEC on May 26, 2010, inter alia, addresses the prudential supervisory disclosure requirement: yearly and half-yearly financial reports "If adopted, this consolidated audit trail would, for the major changes in the holding of voting rights first time ever, allow the SEC and other market regulators ad hoc inside information, which could affect the price to track trade data across multiple markets, products, and of securities. participants in real-time," "It would allow us to rapidly reconstruct trading activity On the other hand, as part of the supervisory disclosure and quickly analyze both suspicious trading behavior and requirements in the EU, according to the Directive 2013/36/ unusual market events." (SEC 2010a, b). EU (Capital Requirements Directive—CRD IV), all EU About the addressed May 6, 2010 crash, the SEC, together member states are required to present information regard- with the CFTC, revealed the full report approximately five ing the laws, regulations, administrative rules and general months later, which was an extremely late action in the Fin- guidance in the field of prudential regulation and supervision Tech world. On May 6, 2010, the U.S. markets suffered not (BaFin&Deutsche Bundesbank 2020). only from an improper algorithm fueled the market crash Publicly held companies and financial services provid- ers, even real person investors in some cases, are under the requirement of public disclosure or supervisory disclosure, To analyze the announcement policies of financial supervisors, we can check the official web sites since supervisors use mainly their web sites to make announcements. Periodic bulletins, press releases, and annual reports are not in the focus of these discussions. The term “Prudential Supervisory Disclosure”, or PSD, was first coined by Zeranski and Sancak (2020). This section mostly depends Algorithmic trading and high-frequency trading are in the domain on their papers. of FinTech. Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 329 but also rumors and detrimental speculations about the driv- ers of the market crash at the time. Prudential supervisory disclosure could save the market from huge material losses and loss of confidence at the time. Prudenal Supervisory It might be less detrimental if some announcements were Supervisory Disclosure by made immediately, and some supervisory actions were taken Disclosure by market during the day. However, many supervisory agencies are not Supervisory parcipants to under the obligation of revealing timely information about Agencies and Supervisory the roots of similar market crashes or events to the public to SROs to the Agencies or Public SROs or the calm down market participants, and markets at large. They Public do this only within their discretion. As in the May 6 case, some supervisors had not been in that capacity, too. There- fore, we see this situation as a new risk for financial markets in our high-speed FinTech world. Public disclosure is mainly a pillar of market discipline and transparency issues. The BCBS touches on supervisory disclosure from the banking sector perspective in the report, Fig. 2 Prudential supervisory disclosure Enhancing Bank Transparency: Public Disclosure and Supervisory Information that Promote Safety and Sound- the markets have all the required information or will have ness in Banking Systems (BIS 1998): “Market discipline, however, can only work if market it timely. The following concerns address the requirement of pru- participants have access to timely and reliable information which enables them to assess a bank’s activities and the dential supervisory disclosure: risks inherent in those activities. Improved public disclosure strengthens market participants’ ability to encourage safe (1) Timing Concern: FinTech has the potential to bring a more radical change within the financial industry and sound banking practices.” We believe that there are multiple drivers for prudential and become a core constituent of its infrastructure and processes, hence boosting the speed and the agility of supervisory disclosure. For example, the global financial cri- sis (2007–2009) has called into question the role of finan- financial services (Kashyap and Weber 2018). In the FinTech world, transactions take place in microsec- cial policy in general, especially in banking, revealing major shortcomings in market discipline, regulation, and supervi- onds. Any late response to the markets might cause severe crashes, FinTech crises and financial crises. sion (The World Bank 2020). Additionally, in the FinTech world, supervisors may have more technological opportuni- (2) Scope Concern: Publicly held companies and FSPs have limited scope, not a market-wide scope. In some ties and potentially strong tools available to carry out their duties. The FinTech world also brings new responsibilities of cases, there might be material information that only supervisors have with their vast information outreach more active market surveillance and more promptly response to the market crashes and abusive market transactions or capacity from multiple data sources. In many regu- latory frameworks, banks, FSPs at large, transmit to news. Moreover, supervisors may have a bigger capacity to contribute to financial stability. From the market quality supervisory authorities, based on a relationship covered by professional secrecy laws and rules, a larger amount perspective, supervisors can contribute more to the informa- tional efficiency of the markets. of accounting data and other information than they are legally required to make public (e.g., annual reports) The structuring a prudential supervisory model is not a complicated work, but it requires a new supervision perspec- or that they publish voluntarily (e.g., in the press) and supervisory authorities can use this important stock of tive. In the private sector, financial services providers have been changing their business models in the FinTech world. information not only to perform the tasks entrusted to them by law but also to enrich the information avail- Their supervisors should also update their business models, or, their supervision perspectives accordingly. able to the public (BIS 1998). As stated with the same BIS report, confidentiality will not be breached if the The following figure shows an interaction between two parties in terms of the prudential supervisory model (Fig. 2). information is released in aggregate forms. In this regard, each supervisory authority should have a predetermined and written prudential supervisory disclo- sure policy. Under this policy, market participants know that 16 –6 One million microseconds are equal to one second (1 µs = 10 s). 330 S. Zeranski, I. E. Sancak (3) Technology Concern: Supervisors potentially have more technological tools and solutions to carry out their duties. SupTech gives supervisors both more Financial Supervisors: Prudenal technological tools and new responsibilities. Comput- Supervisory Disclosure for Their Jurisdicons/Sectors ers can exceed the abilities of human experts in some cases (BaFin 2018). 4) Transparency Concern: Market participants should know under a regulation that supervisors are under the obliga- SROs: Prudenal Suprevisory Disclosure for Their Markets tion of revealing all market-sensitive information, and supervisors will reveal all relevant information to the public. Transparency concern also addresses the infor- Financial Services Providers, Publicly- mational efficiency. traded Companies, Large Shareholders: 5) Mandate Concern: Each financial supervisory authority Supervisory Disclousure/Informaon Disclosure/Public Disclosure has designated mandates and is responsible for success- fully fulfilling the mandates. SupTech applications can turn risk and compliance monitoring from a backward- looking into a predictive and proactive process (Broed- ers and Prenio 2018). Solutions that use advanced data Fig. 3 Laddering PSD analytics and technologies could lead to more timely, dynamic, and even predictive supervision, which enables supervisors to extract knowledge from data that would enable a level playing field for all market participants, and eliminate technology-related asymmetric information risks. be otherwise inaccessible (Dias 2017). Having higher data collection capacity and data analytics tools forces The responsibility of prudential supervisory disclosure might be delegated to SROs when the prudential information supervisors to deliver more useful products and services in a timely fashion. is only in the hand of an SRO. The following figure indicates a possible delegation model (Fig. 3). (6) Accountability Concern: Supervisors should be accountable for their poor disclosure policies to pro- Under the current setup, some supervisors already share their supervisory responsibilities with SROs, such as FINRA tect market integrity and financial consumers as well as financial stability. and exchanges in the U.S. Therefore, a centralized SRO might be assigned as the PSD agency in this regard. Twenty years ago, the regulatory sandbox idea was not A partly similar rule to PSD has already been in prac- tice at the U.S. markets. The Rule 603(b) of Regulation a vision in the supervisory landscape. However, it is well accepted today. Prudential supervisory disclosure is only National Market System (NMS) requires equity exchanges and FINRA to act jointly to disseminate consolidated infor- another concept that we may face soon to handle FinTech related issues properly. mation, including a National Best Bid and Offer (NBBO), on quotations for and transactions in NMS stocks (SEC and To sum up, a SupTech system enables supervisors to collect much better information timely. And, by hav- CFTC 2010). The rule is mainly for fair trading practices. In this regard, the consolidated information is disseminated ing market-wide information, supervisors should reveal information as part of a prudential supervisory disclosure through securities information processors that collect, pro- cess and prepare to publish such information, including the policy to calm down markets, especially in stressful times, increase informational efficiency, cope with market-wide price, size, and symbol of quotations and executions (SEC and CFTC 2010). rumors and increase confidence in the markets. We assume that the initial concern about the PSD model The SEC rules require that the exchanges and FINRA provide timely and accurate data to the Consolidated Tape might be the operational responsibility. It is an acceptable concept that regulators and supervisors should not act as if System (CTS) and Consolidated Quotation System (CQS) to inform all participants of the trading and quoting activi- they are a market actor. They might not be a market actor, but they are also not outside of the markets. Regulators and ties occurring in the market place (SEC and CFTC 2010). According to the SEC and the CFTC’s report (September supervisors and central banks are not entirely outside of the daily market operations. They interfere with markets for the 2010), at the time, there was considerable attention in the public media regarding the data delays, and the staff agreed sake of market integrity and financial stability. On the other hand, staying technological-neutral, with the PSD model, that this was an important topic that should be addressed. PSD is a broader concept and requires informed market par- financial supervisors do not affect the market directions, but contribute to the informational efficiency of the markets, ticipants not only for orders and transactions at one exchange Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 331 or trading center but also for capital market-wide material not obliged to inform financial consumers with all available information that only the supervisory agencies have by their information about banks and banking sector. mandate. After the global financial crisis, bank supervision became Prudential supervisory disclosure is the name of public stricter and more complex, and supervisory capacity did not disclosure for supervisors in the FinTech world. It is a tech- improve proportionally to match the greater complexity of nical requirement with SupTech today. A PSD model might bank regulations (The World Bank 2020). On top of that, the be the idea of flying cars for today, but it seems a reality for SupTech capacity of many financial supervisors could not the future. A statistic says that ninety percent of the data in catch up with the FinTech developments. the world was created in the previous two years alone (IBM As one of the drivers of the global financial crisis, the 2016). Therefore, we should not extrapolate the past too far risk was transferred in nontransparent ways owing to the into the future for technological developments. rapidly increasing trade in complex, structured financial products (The World Bank 2012). Today, some risks might Implications of prudential supervisory disclosure be within the big data, and unless data analytics tools cap- for banking and capital market sectors ture these risks and timely published by the relevant parties to the public, uncertainty might fuel some other FinTech The global financial crisis (2007–2009) has called into crises. Therefore, Know-Your-Data and PSD are crucial for question the role of financial policy in general, especially in financial stability. banking, revealing major shortcomings in market discipline, Using data on publicly traded banks in 61 countries, regulation, and supervision (The World Bank 2020). Fin- Anginer et al. (2018) examined how the institutional envi- Tech also increases the importance of totally new financial ronment affects the relationship between bank capital and policies. The pace of technology, as well as some FinTech system-wide fragility. Their research concludes that bank crashes, addresses the urgency of supervisory reforms. The capital is associated with a reduction in the systemic risk May 6 market crash has many lessons for both capital mar- contribution of individual banks and this effect is more pro- kets and banking sector supervisors. nounced for banks located in countries with less efficient Our analyses mainly focus on the May 6 market events, public and private monitoring of financial institutions and which are about the capital market sector. We prove that the in countries with lower levels of information availability lack of a well-designed SupTech system leaves the capital (Anginer et al. 2018). markets unprotected in the FinTech world. Many economies The study of Demirgüç-Kunt et al. (2008) finds a signifi- still run the same risks today. cant and positive relationship between compliance with the On the other hand, the PSD model also has many implica- Core Principles for Effective Banking Supervision related tions for the banking sector. The decade following the global to information provision and bank soundness. Countries financial crisis was characterized by intense regulation of that require their banks to regularly and accurately report banking sectors worldwide, especially in advanced countries their financial data to regulators and market participants (The World Bank 2020). A decade after the global financial have more highly-rated banks, as timely disclosure of high- crisis, intense regulation seems to be not enough to keep the quality information strengthens monitoring by regulators financial sector safe and sound, since technological develop- and markets alike (Demirgüç-Kunt et al. 2008). Their results ments have been disrupting the sector and transcending the suggest that countries aiming to upgrade banking regulation regulatory issues. and supervision should consider giving priority to informa- The PSD model may be more conducive to the banking tion provision over other elements of the core principles. sector since public disclosure rules and regulations are not similar to the publicly traded companies, which are in the Challenges for financial supervisors realm of the capital market sector. In other words, banks have more regulatory rules to submit information to supervi- In today’s world, in addition to the low pace of digital sors than to financial consumers about their capital require- transformation of supervision with SupTech, one of the ments, operations, organizations, and financial soundness. main challenges is the fragmentation of financial supervi- Therefore, n fi ancial supervisors collect colossal information sion. Supervision is divided among the FED, the FDIC, the from the banking sector, much of them are not available OCC, the SEC, FINRA, the CFTC, and state regulators in for financial sector participants. Moreover, after the global the U.S., centralized bodies such as the ECB, the SSM, the financial crisis, bank regulations became more complex, ESMA, the EIOPA, and the EBA share a stage with national potentially reducing transparency, increasing regulatory arbitrage, and taxing supervisory resources and capacity (The World Bank 2020). However, financial supervisors are We still reserve the privacy issues of banks. A sample of 39 countries. 332 S. Zeranski, I. E. Sancak competent authorities in the EU, and regulation is developed in the next fiscal year. Assuming success in the second fiscal at a national level, and regional coordination is limited in year, two years delay without technology investment makes Asia (Frisell et al. 2018). Regulatory and institutional frame- supervisory agencies old-fashioned in the FinTech world. works will need to be revised in light of new and evolving This picture leaves supervisory agencies unarmed against risks and industry landscapes (Frisell et al. 2018). In other the fast-growing FinTech world. words, as pointed out with a recent paper, Digitalization of What makes the case worse is that operating in a tech- Financial Supervision with Supervisory Technology, before nologically leading country does not help the U.S. financial the digital transformation, a check-up for the whole financial supervisors carry out their responsibilities successfully, but system and adjustments for the financial structure are the the technology stirs mostly financial markets and institu- prerequisites of having a modern and functional supervisory tions, namely the private sector. This legacy political struc- system. ture is per se a source of risk for financial stability. As we The regulatory and supervisory framework in the EU observe that the major financial reforms have been followed does not directly address the RegTech or SupTech para- mostly by crises or scandals, unless politicians do not get digms, and the approach taken by firms and supervisors to pressure from lobbying channels, they are not inclined to pilot and adopt RegTech and SupTech frameworks is cur- increase the budget for financial supervisors. This situation rently ad-hoc and uncoordinated (European Commission also seems a kind of vicious circle or dilemma for the stabil- 2019). This was seen as an important issue and handled by ity of the financial industry and the global economy at large. a report, Expert Group on Regulatory Obstacles to Financial Innovation (ROFIEG):30 Recommendations on Regulation, Innovation and Finance -Final Report to the European Com- Conclusion mission. The Group recommends that the EU develops and implements a comprehensive and ambitious agenda for the Financial supervisors collect vast amounts of data and infor- establishment of advanced RegTech and SupTech capabili- mation about market institutions and activities, but they are ties, in coordination with relevant authorities in and beyond not obliged to reveal information under a specific disclosure the EU and international standard setters. policy. They inform the public within their discretion but It is estimated that finance goes real-time, and periodic not under a predetermined disclosure policy. This should reporting no longer drives operations and decisions in the not be the case anymore in the FinTech world with a SupT- near future (Deloitte 2020). And, speed has always been of ech capacity. As financial institutions have the responsibil- the essence in financial markets (Ait-Sahalia and Saglam ity of Know-Your-Customer, financial supervisors should 2013). Therefore, supervisors also should equip with high- also have “Know-Your-Markets”, “Know-Your-Technology”, speed technological tools to respond to market crashes and “Know-Your-Data”, and “Inform-Your-Markets” responsi- FinTech crises properly. bilities. With the PSD model, from the market quality and FinTech players often fall outside the applicable regu- micro-structure perspectives, supervisors can contribute latory and supervisory framework both for prudential and more to the informational efficiency of the markets. customer protection supervisions, which is a challenge that The U.S. financial markets faced an unprecedented rapid regulators or supervisors with capacity constraints may be decline and recovery on May 6, 2010, known as the May ill-equipped to address (Berg et al. 2020). The U.S. finan- 6 flash crash. Roughly one trillion $ market value in less cial markets are the most complex markets from supervisory than thirty minutes vanished with the biggest one-day point perspectives. The magnitude of transactions, fragmented decline in the history of the DJIA at the time. Since the but interconnected markets, the variety of financial instru - market events took place in electronic markets, and algo- ments coupled with the fragmented and intricate design of rithmic trading (AT) and high-frequency trading (HFT), the financial regulatory and supervisory structure make the parts of FinTech, played significant roles, we handle the markets incredibly difficult to monitor, manage FinTech May 6 flash crash from the FinTech, SupTech, and financial related risks, and cope with financial frauds. supervision perspectives. Our research is unique because Even though financial supervisors are independent in we analyzed the May 6, 2010 flash crash first time from their responsibility areas, their budgets are tied to political FinTech and SupTech, or “TECHs in Finance” perspectives. decisions. In many countries, budget allocations to inde- We flashbacked the events and analyzed the responses of the pendent financial agencies take place only once a year. Since economic and financial media and two U.S. financial super - the financial sector of the U.S. is extraordinarily complex, visors, the SEC and the CFTC, to the market events. The modernization and updating works require not millions but case has many lessons and takeaways for the governments, billions of U.S. dollars. If an agency cannot get a budget economic policymakers, and regulatory and supervisory increase for technology investments or reform requirements authorities, and academic communities. Our analyses are for a fiscal year, then the agency has the chance to get it only Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 333 more conducive to the U.S. and the EU because their frag- PSD capacity, the EU Member States, the U.S., and many mented financial systems are in the urgent need of TECHs other countries run similar risks. in Finance reforms. Analyzing the May 6 flash crash, we find that the techno- Funding Open Access funding enabled and organized by Projekt logical imbalance between financial markets or institutions DEAL. and their supervisors drove the markets in uncertainty, hence in a fear and panic environment. Since the imbalance has not Compliance with ethical standards diminished yet, the same risks still exist. As a remedy, we introduce a new concept, prudential supervisory disclosure Conflict of interest The authors certify that they are not affiliated with (PSD), and a model, the PSD model, with a well-function- or involved in any organization or entity with any financial interest or ing SupTech system, to cope with the May 6 type FinTech non-financial interest in the subject matter or materials discussed in this manuscript. crises. Even though the U.S. has been one of the leading technology innovating countries in the world, the May 6 Open Access This article is licensed under a Creative Commons Attri- case indicated that the U.S. financial supervisors, namely the bution 4.0 International License, which permits use, sharing, adapta- SEC and CFTC did not have good enough SupTech capac- tion, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, ity at the time. We are convinced that it is not about having provide a link to the Creative Commons licence, and indicate if changes available technology; it is about organizing, designing, and were made. The images or other third party material in this article are having a well-functioning SupTech and supervisory system included in the article’s Creative Commons licence, unless indicated at large. Moreover, having a full-fledged SupTech system is otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not not wholly tied to financial supervisors. It requires additional permitted by statutory regulation or exceeds the permitted use, you will funds and hence political support as well as leadership. need to obtain permission directly from the copyright holder. To view a Risk management policies were developed mostly after a copy of this licence, visit http://creativ ecommons .or g/licenses/b y/4.0/. crisis comes out in the financial sector. However, we do not have such a comfortable reform approach any more against FinTech crises. Due to the nature of such crises, markets and institutions can be wiped out in hours, if not in minutes. References The May 6 market crash depleted market liquidity in twenty minutes, collapsed prices, and caused a massive panic at the Ait-Sahalia, Y., and M. Saglam. 2013. High Frequency Traders: Taking U.S. financial markets. Advantage of Speed. Cambridge: National Bureau of Economic Research. PSD is the name of public disclosure for supervisors in Anginer, D., A. Demirgüç-Kunt, and S.D. Mare. 2018. Bank capital, the FinTech world. It is a technical requirement with SupT- institutional environment and systemic stability. Journal of Finan- ech today. The PSD model helps protect market integrity by cial Stability, pp. 97–106. revealing useful information timely about market functions Arner, D.W., J. Barberis, and R.P. Buckley. 2016. The emergence of RegTech 2.0: from know your customer to know your Data. Jour- and against improper market activities or rumors. A PSD nal of Financial Transformation, pp. 79–86. model might be the concept of flying cars for today, but it BaFin, F. 2018. Big Data Meets Artificial Intelligence: Challenges and seems a reality for the future. Taking seriously a statistic Implications for the Supervision and Regulation of Financial Ser- about 2017 trends saying that ninety percent of the data in vices. Bonn: BaFin. B.D. Bundesbank. 2020. Supervisory Disclosure. Retrieved from the world was created in the previous two years alone, we do Supervisory Disclosure: https ://www.super visor y-discl osure .de/ not extrapolate the past too far into the future for technologi- super visor y-en/ cal developments. Thus, we expect the PSD model or a ver- Berg, G., M. Guadamillas, H. Natarajan, and A. Sarkar. 2020. Fintech sion of the model as the next normal of the financial sector. in Europe and Central Asia: Maximizing Benefits and Managing Risks. Washington, DC.: The World Bank Group. FinTech crises might cause bank runs and destroy the BIS. 1998. Enhancing Bank Transparency: Public Disclosure and baking sector as well as capital markets. One of the initial Supervisory Information that Promote Safety and Soundness in considerations for the May 6 market crash was about fat Banking Systems. Basle: Bank for International Settlements-Basle finger speculations for a bank. This makes the case more Committee on Banking Supervision. Borch, C. 2017. High-frequency trading, algorithmic finance, and the important; banks are vulnerable to rumors which might trig- flash crash: reflections on eventalization. Economy and Society, ger bank runs. In this regard, the PSD model is crucially pp. 350–378. important to protect banks from FinTech crises and bank Bouveret, A., P. Breuer, Y. Chen, D. Jones, and T. Sasaki. 2015. Fra- runs. We contemplate that the May 6 case could have been gilities in the U.S. Treasury Market: Lessons from the “Flash Rally” of October 15, 2014. Washington. D.C.: International more detrimentally and driven the banks into collapse under Monetary Fund. the supervisory setup at the time. And, it could have been Broeders, D., and J. Prenio. 2018. Innovative Technology in Financial less detrimental under a PSD model. Currently, without a Supervision (Suptech)-the Experience of Early Users. Basel: Bank for International Settlements-Financial Stability Institute. 334 S. Zeranski, I. E. Sancak Bullock, N., and P. Stafford. 2019. SEC Approves Vast Surveillance Kirilenko, A.A., A.S. Kyle, M. Samadi, and T. Tuzun. 2017. The flash System for Stock Market. Retrieved from Financial Times: https: // crash: high-frequency trading in an electronic market. Journal of www.ft.com/conte nt/2a428 156-ab1e-11e6-9cb3-bb820 79021 22. Finance, pp. 967–998. CFTC and SEC. (2010). Preliminary Findings Regarding the Market Kirilenko, A.A., A.S. Kyle, M. Samadi, and T. Tuzun. 2018. Automa- Events of May 6, 2010. washington, D.C.: CFTC and SEC. tion, intermediation and the flash crash. Journal of Investment CNBC. (2010). FLASH CRASH May 6, 2010 (Part 6 of 6) CNBC. Management, pp. 29–46. Retrieved from Youtube: https ://www.y outu be.com/w atc h Krantz, M. (2014). Four-year flash crash anniversary haunts markets. ?v=7UhKO s3dYk 4. Retrieved from CNBC: https ://www.cnbc.com/2014/05/05/four- CNBC. 2014. The Lasting Impact of the 2010 Flash Crash. Retrieved year-flash -crash -anniv ersar y-haunt s-marke ts.html. from CNBC: https ://www .cnbc.com/2014/05/06/the-las ti ng-im pac Krudy, E. (2010). Stocks plunge as trading glitch suspected. Retrieved t-of-the-2010-flash -crash .html. from Reuters.com: https ://www.reute rs.com/artic le/us-marke ts- Deloitte, J.A. 2020. Digital Finance Series: Finance 2025. London: stock s/stock s-plung e-as-tradi ng-glitc h-suspe cted-idUST RE634 Deloitte.1EA20 10050 6. Demirgüç-Kunt, A., E. Detragiache, and T. Tressel. 2008. Banking on Lauricella, T., and P.A. McKay. 2010. Dow Takes a Harrowing the principles: compliance with Basel core principles and bank 1,010.14-Point Trip; Biggest Point Fall, Before a Snapback; Glitch soundness. Journal of Financial Intermediation, pp. 511–542. Makes Thing Worse. Retrieved from WSJ: https ://www.wsj.com/ Dias, D. 2017. FinTech, RegTech and SupTech: What They Mean for artic les/SB100 01424 05274 87043 70704 57522 77541 31412 596. Financial Supervision. Toronto: Toronto Centre. Lexico. 2020. Definition of prudential in English. Retrieved from Lex- Easley, D., M. Lopez de Prado, and M. O’Hara. 2011. The microstruc- ico: https ://www.lexic o.com/en/defin ition /prude ntial . ture of the ‘Flash Crash’: flow toxicity, liquidity crashes and the Mackenzie, M. 2010. Vital lessons of the ‘flash crash’. Retrieved probability of informed trading. The Journal of Portfolio Manage- from ft.com/: https ://www.ft.com/conte nt/2e1a9 6ee-5f7e-11df- ment, pp. 118–128.a670-00144 feab4 9a. EBA. 2018. The EBA’s FinTech Roadmap: Concluions from the Consul- Menkveld, A. J., and B.Z. Yueshen. 2018. The flash crash: a caution- tation on the EBA’s Approach to Financial Technology (FinTech). ary tale about highly fragmented markets. Management Science, Paris: EBA. pp. 4470–4488. EBA. 2020. EBA Response: EC consultation on the Digital Finance Mishkin, F. S. 2000. Prudential Supervision: What Works and What Strategy/Action Plan. Paris: EBA. Doesn’t? NBER Conference-Working Paper 7926 (pp. 1–36). European Central Bank. 2017. A Quick Guide to Macroprudential Poli- Florida: National Bureau of Economic Research. cies. Retrieved from European Central Bank: https ://www.ecb. Orol, R.D. 2010. SEC: No evidence that one event caused market eur op a.eu/expla iners /tell-me-mor e/html/macr o pr ude ntial polic plunge. Retrieved from Marketwatch.com: https ://www .mar k e ies.en.html.twatch.com/s tory/no-e vidence-of-f at-finger -error-in-ma y-6-plung European Commission. 2020. Prudential Requirements. Retrieved from e-2010-05-11. European Commission: https ://ec.europ a.eu/info/busin ess-econo Peress, J. 2014. The Media and the Diffusion of Information şn Finan- my -eur o/banki ng-and-f inan ce/f inan cial-super visio n-and-r isk - cial Markets: evidence from Newspaper Strikes. The Journal of manag ement /manag ing-r isk s -bank s -and-finan cial-ins ti tutio ns/ Finance, pp. 2007–2043. prude ntial -requi remen ts_en. Podkul, C. 2020. When Volatility Surges, SEC’s Trade-Monitoring Sys- European Commission. 2019. Expert Group on Regulatory Obstacles tem Has Struggled. Retrieved from WSJ: https ://www.wsj.com/ to Financial Innovation (ROFIEG):30 Recommendations on Reg-articles/when-v olatility -sur ges-secs-tr ade-monit or ing-sy stem-has- ulation, Innovation and Finance -Final Report to the European strug gled-11585 79644 2?mod=searc hresu lts&page=1&pos=5. Commission—December 2019. Brussel: European Commission. Schapiro, M. L. 2010. Testimony Concerning the Severe Market Dis- European Commission. 2020. Transparency Requirements for Listed ruption on May 6, 2010. Retrieved from SEC: https ://www.sec. Companies. Retrieved from European Commission: https ://gov/news/testi mony/2010/ts051 110ml s.htm. e c. e ur o p a .e u /i n fo /b u si n e ss - e co n o my- eu r o/ c om pa ny- r e p or t in g - Schapiro, M.L. 2010. Examining the Causes and Lessons of the and-auditing/com pan y-reporting/tr anspar ency -requir ements-lis te May 6th Market Plunge. Washington, D.C.: U.S. Securities and d-compa nies_en. Exchange Commission. European Parliament. 2018. Report on the proposal for a directive SEC. 2010. SEC Proposes Consolidated Audit Trail System to Better of the European Parliament and of the Council on the pruden- Track Market Trades. Retrieved from SEC: https ://www.sec.gov/ tial supervision of investment firms and amending Directives news/press /2010/2010-86.htm. 2013/36/EU and 2014/65/EU. Retrieved from https:/ /www.europ SEC. 2012. Investor Bulletin: Measures to Address Market Volatil- ar l.eur op a.eu/: https ://www .eur op ar l.eur op a.eu/doceo /docum ity. Retrieved from SEC: https ://www.sec.gov/oiea/inves tor-alert ent/A-8-2018-0295_EN.pdf.s-bulle tins/inves tor-alert s-circu itbre akers bulle tinht m.html. FINRA. 2020. catnmsplan. Retrieved from FINRA CAT: https://www . SEC. 2013. MIDAS-Market Information Data Analytics System. catnm splan .com/. Retrieved from SEC: https://www .sec.gov/markets tructur e/midas Frisell, L., M.J. Fernandes, L. Quest, E. Rennick, S. Roy, and D. .html#.XoTYl 4hKg_E?mod=artic le_inlin e. Treeck. 2018. Supervising Tomorrow. New York: Oliver Wyman. SEC. 2019. SEC Approves Plan to Create Consolidated Audit Trail. Gomber, P., J.A. Koch, and M. Siering. 2017. Digital finance and Fin- Retrieved from U.S. Securities and Exchange Commission: https Tech: current research and future research directions. Journal of ://www.sec.gov/news/press relea se/2016-240.html. Business Economics, pp. 537–580. SEC. 2020. About EDGAR. R etrieved from SEC: https://www .sec.gov/ IBM. 2016. 10 Key Marketing Trends for 2017. New York: IBM.edgar /about . Janjigian, V. 2010. Fat Fingers Cause Panics. Retrieved from Forbes. SEC. 2020. SEC Proposes Improvements to Governance of Market com: https://www .forbes.com/sites /g reatspecu latio ns/2010/05/06/ Data Plans. Retrieved from SEC: https: //www.sec.gov/news/press fat-finge rs-cause -panic s/#180ad fe160 e9.-relea se/2020-5. Kashyap, K., and M.G. Weber. 2018. How Emerging Technologies will SEC and CFTC. 2010. Findings Regarding the Market Events of May Change Financial Services. In The FinTech Book, ed. S. Chisti and 6, 2010: Report of the Staffs of the CFTC and SEC to the Joint J. Barberis, 228. Cornwall: Wiley. Advisory Committee on Emerging Regulatory Issues. Washington, D.C.: SEC and CFTC. Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons… 335 SEC. 2010. Concept Release: Concept Release on Equity Market Struc-://www.frbat lanta .org/cenfi s/publi catio ns/notes fromt hevau lt/11- ture; Proposed Rule. Retrieved from SEC: https ://www.sec.gov/prudential -r egulation -bigda t a-and-machine-lear ning-2016-11-21. rules /conce pt/2010/34-61358 fr.pdf. Zeranski, S., and I.E. Sancak. 2020. Digitalisation of Financial Super- The World Bank. 2012. Rethinking the Role of the State in Finance: vision with Supervisory Technology (SupTech). Journal of Inter- Global Financial Development Report 2013. Washington, D.C.: national Banking Law and Regulation, pp. 309–330. The World Bank. The World Bank. 2020. Bank Regulation and Supervision a Decade Publisher’s Note Springer Nature remains neutral with regard to after the Global Financial Crisis: Global Financial Development jurisdictional claims in published maps and institutional affiliations. Report 2019–2020. Washington, D.C.: The World Bank. Wall, L.D. 2016. Prudential Regulation, Big Data, and Machine Learning. Retrieved from Federal Reserve Bank of Atlanta: https

Journal

International Journal of Disclosure and Governance – Springer Journals

Published: Dec 1, 2021

Keywords: Supervisory technology; SupTech; FinTech; RegTech; Financial supervision; Financial system; Prudential supervisory disclosure; Financial authority; Digital finance; May 6 flash crash; FinTech crises; Financial crises; Financial stability; Informational efficiency; Systemically important data; Systemically important markets; Know-your-data; Know-your-technology; Know-your-markets; Inform-your-markets; D47; D53; G18; G01; G28; H11; K22; K23; L15; O31; O32

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons from a FinTech crisis

Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons from a FinTech crisis

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons from a FinTech crisis

Prudential supervisory disclosure (PSD) with supervisory technology (SupTech): lessons from a FinTech crisis

References (56)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies